THE EFFECTS OF MULTIPLE HANDLERS ON DEVELOPMENT OF DNA PROFILES FROM ASSEMBLERS OF DEFLAGRATED PIPE BOMBS By Jade McDaniel A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Forensic Science 2012 ABSTRACT THE EFFECTS OF MULTIPLE HANDLERS ON DEVELOPMENT OF DNA PROFILES FROM ASSEMBLERS OF DEFLAGRATED PIPE BOMBS By Jade McDaniel The use of improvised explosive devices (IEDs) in terrorist attacks has become a more common occurrence in the United States. The increase in usage has made it imperative to find methods for identifying manufacturers of IEDs. Previous research has focused on methods for recovering DNA from deflagrated pipe bombs and obtaining an STR profile of the individual who assembled the device. These studies all examined scenarios where a single individual handled a pipe bomb that had previously been decontaminated to rid it of any preexisting DNA. Although such controlled studies are informative, in actuality the components of a pipe bomb could be handled by other people before the assembler takes possession of them. The goal of this research was to assess how multiple handlers affect the development of DNA profiles from the assemblers of pipe bombs following deflagration. Thirty-four volunteers mock assembled pipe bombs that had not been previously decontaminated. Bombs were deflagrated and the fragments collected and swabbed. Recovered DNA was quantified, amplified, and typed using short tandem repeat (STR) analysis. There was a high level of mixed profiles, as over half of the samples were classified as such. In spite of this, profiles consistent with the handler were developed for 11 of the 34 bombs. These results indicate that it is possible to obtain handler profiles from deflagrated pipe bomb fragments, even when bomb components have not undergone any prior cleaning procedures. ACKNOWLEDGEMENTS I would like to acknowledge and extend my gratitude to those who made this research possible. Thank you to Dr. Ruth Smith and Dr. David Foran for their time spent editing this thesis. Special thanks go to SPL/SGT Timothy Ketvirtis and the other members of the Michigan State Police Bomb Squad for their assistance throughout the entire deflagration process. Also, thanks goes to Mr. William Nelson of Operating Engineers Local 324 for use of his facilities. Thank you to Dr. Thomas Holt, for serving as one of my committee members. My thanks extend to my classmates who assisted in this project. Thank you to Stephen Gicale, for the insight he provided me throughout this research. Thank you to Keiauda Tennant and Alicya Orlando for helping me design and orchestrate the volunteer handling sessions. Thank you to all other members of the Michigan State University Forensic Biology group for their support. Also, thanks go to the volunteers who handled the pipe bombs. Finally, I would like to extend my gratitude to my family and friends for their support throughout my academic career. I greatly appreciate all the insight and advice they have given me. A portion of this research was sponsored through a contract entitled Genetic Identification of the Manufacturers of Improvised Explosive Devices, issued by the Technical Support Working Group to David Foran PhD and Michigan State University. All points of view in this manuscript are solely those of the author. iii TABLE OF CONTENTS List of Tables ...................................................................................................................................v List of Figures ............................................................................................................................... vii Introduction ......................................................................................................................................1 Analysis of low copy number DNA .........................................................................................3 Interpretation of DNA mixtures ...............................................................................................6 Previous studies focusing on the recovery of DNA from deflagrated pipe bombs ..................7 Goals of the current study.........................................................................................................9 Materials and methods ...................................................................................................................10 Preliminary analysis of the effects of multiple handlers using end caps and pipe bombs .....10 Pipe bomb preparation and handling ......................................................................................12 Pipe bomb deflagration and collection ...................................................................................13 DNA recovery and extraction from deflagrated pipe bombs .................................................14 DNA quantification, amplification, and STR analysis ...........................................................17 Processing of reference samples .............................................................................................18 Development of consensus profiles ........................................................................................18 Statistical analyses ..................................................................................................................19 Results ............................................................................................................................................21 Preliminary analysis of the effects of multiple handlers using end caps and pipe bombs .....21 Pipe bomb deflagration...........................................................................................................22 DNA recovery and extraction from deflagrated pipe bombs .................................................23 DNA quantities recovered from deflagrated pipe bombs .......................................................24 Analysis of drop-out based on locus ......................................................................................25 Categorizing mixtures based on number of loci containing extra alleles ...............................26 Categorizing loci within consensus profiles based on the presence of handler alleles ..........27 Categorizing consensus profiles based on number of loci containing handler’s alleles ........28 Consensus profile quality compared to DNA recovery..........................................................29 Analysis of allele calls in consensus profiles that were inconsistent with the handler’s .......30 Discussion ......................................................................................................................................32 Conclusion .....................................................................................................................................42 Appendix A. DNA quantities recovered from pipe bombs............................................................43 Appendix B. Allele calls and loci categorizations .........................................................................47 References ......................................................................................................................................70 iv LIST OF TABLES Table 1: An example of swab labeling using pipe bomb #27 ........................................................16 Table 2: Consensus profiles compared to last handler profiles in preliminary studies .................22 Table 3: Mann-Whitney Rank Sum Test for comparison of DNA recovered from end caps and pipes .................................................................................................................................25 Table 4: Comparison of consensus profiles quality and average DNA recovery ..........................30 Table 5: Mann-Whitney Rank Sum Test for comparison of DNA recovered within each consensus profile category ................................................................................................30 Table 6: DNA quantities recovered from pipe bombs ...........................................................45 – 46 Table 7: Allele calls and loci categorizations for pipe bomb 1......................................................49 Table 8: Allele calls and loci categorizations for pipe bomb 2......................................................50 Table 9: Allele calls and loci categorizations for pipe bomb 3......................................................51 Table 10: Allele calls and loci categorizations for pipe bomb 4....................................................51 Table 11: Allele calls and loci categorizations for pipe bomb 5....................................................52 Table 12: Allele calls and loci categorizations for pipe bomb 6....................................................52 Table 13: Allele calls and loci categorizations for pipe bomb 7....................................................53 Table 14: Allele calls and loci categorizations for pipe bomb 8....................................................53 Table 15: Allele calls and loci categorizations for pipe bomb 9....................................................54 Table 16: Allele calls and loci categorizations for pipe bomb 10..................................................54 Table 17: Allele calls and loci categorizations for pipe bomb 11..................................................55 Table 18: Allele calls and loci categorizations for pipe bomb 12..................................................55 Table 19: Allele calls and loci categorizations for pipe bomb 13..................................................56 Table 20: Allele calls and loci categorizations for pipe bomb 14..................................................56 v Table 21: Allele calls and loci categorizations for pipe bomb 15..................................................57 Table 22: Allele calls and loci categorizations for pipe bomb 16..................................................57 Table 23: Allele calls and loci categorizations for pipe bomb 17..................................................58 Table 24: Allele calls and loci categorizations for pipe bomb 18..................................................58 Table 25: Allele calls and loci categorizations for pipe bomb 19..................................................59 Table 26: Allele calls and loci categorizations for pipe bomb 20..................................................60 Table 27: Allele calls and loci categorizations for pipe bomb 21..................................................60 Table 28: Allele calls and loci categorizations for pipe bomb 22..................................................61 Table 29: Allele calls and loci categorizations for pipe bomb 23..................................................61 Table 30: Allele calls and loci categorizations for pipe bomb 24..................................................62 Table 31: Allele calls and loci categorizations for pipe bomb 25..................................................63 Table 32: Allele calls and loci categorizations for pipe bomb 26..................................................63 Table 33: Allele calls and loci categorizations for pipe bomb 27..................................................64 Table 34: Allele calls and loci categorizations for pipe bomb 28..................................................65 Table 35: Allele calls and loci categorizations for pipe bomb 29..................................................66 Table 36: Allele calls and loci categorizations for pipe bomb 30..................................................66 Table 37: Allele calls and loci categorizations for pipe bomb 31..................................................67 Table 38: Allele calls and loci categorizations for pipe bomb 32..................................................67 Table 39: Allele calls and loci categorizations for pipe bomb 33..................................................68 Table 40: Allele calls and loci categorizations for pipe bomb 34..................................................68 Table 41: Allele calls for pipe bomb RB1 .....................................................................................69 Table 40: Allele calls for pipe bomb RB2 .....................................................................................69 vi LIST OF FIGURES Figure 1: Steel pipe bomb ............................................................................................................. 13 Figure 2: Cement structure ........................................................................................................... 14 Figure 3: Metal crate used to retain pipe bomb fragments ........................................................... 14 Figure 4: Deflagrated pipe bomb separated into six sections ....................................................... 15 Figure 5: Pipe bomb fragmentation .............................................................................................. 23 Figure 6: Characterization of drop-out based on locus ................................................................. 26 Figure 7: Categorization of mixtures based on number of loci with extra alleles ........................ 27 Figure 8: Locus categorizations based on the presence of handler alleles ................................... 28 Figure 9: Consensus profile categorizations based on number of loci with handler's alleles ....... 29 vii INTRODUCTION Since the early 1990’s terrorist attacks targeting the United States have become a more common occurrence (FBI Terrorist Report 2000 – 2005). The purpose of such attacks began as destruction of facilities, equipment, supplies, and resources; however, mass casualty to the populace is now the primary objective. This change in objective has prompted the development of new methods to produce the deadliest results. Currently, improvised explosive devices (IEDs) are the most commonly used weapons in terrorist attacks (DOD, 2006; Smith, 2011). IEDs are ‘homemade’ devices fabricated from military or commercially available explosive components, that when assembled, are designed to destroy, maim, or wound (DOD, 2006). Recent attacks against US military forces in Iraq and Afghanistan have been carried out by unofficial paramilitary units whose mission is terrorism, rebellion, and guerilla warfare (Belmont et al., 2010). As a result of this insurgency, most American combat casualties currently result from IEDs (Ramasamy et al., 2008; Belmont et al., 2010). Between June 2003 and March 2011 over 1,800 soldiers in Iraq, and more than 1,000 soldiers in Afghanistan were killed by IEDs (iCasualties.org, March 2011). In 2010 alone, there were 368 IED fatalities in Afghanistan (iCasualties.org, March 2011). Owens et al. (2008) conducted a study addressing the wounding patterns and mechanisms in the Iraq and Afghanistan conflicts and found that there are a greater number of casualties being caused by explosions than by bullets. Although IEDs exist in many different forms, the use of steel pipe bombs has become increasingly common among terrorists in the United States (Gibbons et al., 2003). IEDs are popular because they can be constructed easily from materials that are commonly available (Gibbons et al., 2003; DOD, 2006). For example, the simplest of IEDs require only a piece of 1 pipe, two end caps, and explosive powder, all of which are inexpensive and easily obtained from a sporting goods store (Gibbons et al., 2003). The explosive is loaded into the pipe, which is sealed by the end caps, and a triggering mechanism, often a simple fuse, is installed (Gibbons et al., 2003; DOD, 2006). The explosion of a pipe bomb causes bomb fragments and shrapnel to be propelled at high speeds, resulting in bodily harm and/or damage to structures (Gibbons et al., 2003). Another appealing aspect of a pipe bomb is that it can be placed at the scene and detonated from a distance, providing safety for the perpetrator (DOD, 2006). The identity of the perpetrator then becomes very difficult to determine. Despite the destruction caused by a bombing, remaining evidence can contain valuable information that may help identify the individual(s) who constructed the bomb. The FBI Handbook of Forensic Services (2007) outlines the procedure for collecting evidence after an explosion. During examination of a bomb scene, investigators attempt to identify any components used in making the bomb, such as fragments of pipe, switches, detonators, or wires. The examination of components also helps to determine the intended function of the device or any unique techniques used by the assembler during its construction. Analyzing the physical characteristics of the individual components of the device allows them to be linked to stores or factories where they may have been sold or produced, which in turn could link them to a potential suspect. Once collected, bomb fragments are examined for the presence of explosives residue. This residue can be analyzed to identify the specific type of explosive that was used and if it is similar to any other explosives that have been previously used in similar devices. It is also important to preserve any biological material, such as fingerprints or DNA that have been deposited on the components, as this material could provide additional information about the identity of the assembler (FBI Handbook of Forensic Services, 2007). 2 The increase in IED usage in Iraq and Afghanistan has prompted research focused on the detection and prevention of IED threats. Although prevention of any attacks would be ideal, it is not always possible. Therefore, there has been a shift in research efforts to explore methods that will aid in the identification of the bomb assembler(s) after an attack has taken place (Smith, 2011). Recent studies have demonstrated that DNA can be recovered from deflagrated pipe bomb fragments (Esslinger et al., 2004; Kremer, 2008; Bille et al., 2009; Foran et al., 2009; Gomez, 2009). However, the quality and quantity of DNA that is recovered from deflagrated pipe bombs will likely be poor, making it difficult to obtain quality data that can be reliably used to identify suspects. Furthermore, several individuals may have had contact with the bomb components; consequently any DNA that is recovered may be cross-contaminated by multiple contributors. These factors have the potential to affect DNA analysis and interpretation, thus methods need to be developed that can test the feasibility of matching one profile, among several potential profiles deposited on the bomb fragments, to the individual(s) who assembled that particular bomb. Analysis of Low Copy Number DNA When a person comes into contact with a surface, skin cells may be sloughed off and left on it. The DNA within these skin cells has the potential to be analyzed in a procedure commonly referred to as ‘touch DNA’ analysis (Daly et al., 2011). The quantity of DNA left behind when an individual touches an object is variable, and can depend upon several factors including the duration of the interaction between the person and the object, the material from which the object is made, and the tendency of an individual to ‘shed’ epithelial cells (Farmen et al., 2008; Goray et al., 2010). It has been shown that full genetic profiles can be obtained from touch DNA 3 retrieved from items such as test tubes, glass vials, and cloth which then has been used successfully to identify the person who came in contact with that object (Phipps and Petricevic, 2007; Daly et al., 2011). In cases where the period of contact is brief or the person does not easily deposit skin cells upon contact, the amount of DNA obtained from the item may be defined as low copy number (LCN), which refers to instances when quantities of DNA below 100 – 200 pg are used during amplification (Gill and Buckleton, 2010). These low levels of DNA are often not sufficient to provide reliable results because they fall below the established instrument thresholds for detection (Gill and Buckleton, 2010). Forensic laboratories generally use segments of DNA, called short tandem repeats (STRs), to assemble a DNA profile from biological evidence. STRs are highly variable among individuals, making them ideal genetic loci for creating a DNA profile that is unique to an individual (Butler, 2006). Commercially available STR amplification kits require approximately 1 ng of DNA in order to yield consistent, reliable DNA profiles (Applied Biosystems, 2007). Quantities less than this produce DNA profiles that can be affected by one or more stochastic (random) effects that interfere with the correct identification of genotypes. Examples include increased heterozygote peak imbalance, allelic drop-out/drop-in, or increased stutter peaks (Balding and Buckleton, 2009; Cowen et al., 2010). In general, heterozygous alleles should have peak heights that are greater than 70% of the highest allele in the pair; however, when dealing with LCN DNA, heterozygote peak imbalance is often exacerbated because one allele can be under-represented when sampled. At the extreme this can result in complete allelic drop out, which is when an allele fails to amplify. Allelic drop-in refers to additional peaks within a profile. Stutter peaks are caused by ‘slippage’ of the DNA polymerase during replication of the 4 DNA strand, which results in a peak that is one repeat unit smaller or larger than the corresponding true allele. In many cases only very small amounts of DNA are available from evidence to assemble a DNA profile to solve a case. This has also shown to be true for DNA recovered from pipe bomb fragments (Esslinger et al., 2004; Gomez, 2009). Due to this, efforts have been made to establish new, or modify current, procedures to obtain DNA profiles from such small quantities of DNA. Among these are increasing the number of amplification cycles to increase product yield, reducing the polymerase chain reaction (PCR) volume to increase the concentration of the amplified product, cleaning up the amplified sample to remove primers or salts that compete with the injection of DNA into the capillary, and increasing injection time during electrophoresis which allows more DNA to be injected into the capillary (Balding and Buckleton, 2009). The development of modified short tandem repeat (STR) kits, known as miniSTRs, increases the likelihood of obtaining a profile from degraded, LCN DNA (Coble and Butler, 2005). In miniSTRs, the primers bind closer to the region of interest, generating a smaller region of DNA to be amplified. During LCN DNA analysis, even after protocol modifications have been made, it has been recommended that replicate analyses be performed (Balding and Buckleton, 2009). Replicates allow an allele to potentially be observed multiple times, thus increasing confidence in that particular allele call. If an allele is not observed at least twice in replicate samples, then it is not considered to be real. Previous research employing the consensus method, which is based on replicate analyses, has shown this technique to be effective in overcoming the stochastic effects commonly observed in LCN DNA analysis (Cowen et al., 2010). 5 Interpretation of DNA Mixtures In addition to the difficulties associated with obtaining profiles from LCN DNA, evidentiary items can also be contaminated with DNA from multiple individuals. Mixtures are extracts in which DNA from multiple individuals has been deposited, co-extracted, and amplified, thus complicating interpretation of the results (Schneider et al., 2009). The most obvious indication of a mixed profile is the presence of three or more allele peaks at two or more loci although, there are various scenarios that can result in extra peaks within a DNA profile (Schneider et al., 2009). For instance, multiple peaks can result from stutter. Typically, stutter peaks are less than 15% of the height of their corresponding allele peak, however when working with LCN DNA the height of the stutter peak can be larger than 15%, causing the appearance of a mixed profile, making interpretation difficult (Budowle et al., 2009). Another possible cause of additional peaks is the incomplete addition of an adenine to the terminus of a newly synthesized DNA strand. If the nucleotide does not get added it can result in two allele peaks that are separated by one base pair, although both peaks represent the same allele. Random dropin can also cause extraneous peaks. The interpretation of a mixed profile is further complicated because it depends on the ratio of DNA present from each individual (Budowle et al., 2009). Some mixtures may have equal proportions of each contributor, while others may have unequal proportions. As a result of the inter-individual variation in the amount of skin cells shed by an individual, it is possible that not all individuals represented on the evidence will yield a complete DNA profile. Thus, it is often difficult to determine the number of individuals contributing to a mixture. Due to the complications surrounding mixture interpretation, researchers have proposed guidelines to aid in their analysis. These guidelines are comprised of six main steps, which are to (1) identify the 6 presence of a mixture, (2) designate true allele peaks, (3) identify the number of possible contributors, (4) estimate the mixture proportions, (5) consider all possible genotype combinations, and (6) compare possible profiles to reference samples (Gill et al., 2006; Budowle et al., 2009). Software programs, such as Linear Mixture Analysis (Perlin and Szabady, 2001) and PENDULUM (Bill et al., 2005) have also been developed to aid in mixture interpretation. In general, these are based on mathematical approaches to estimate possible mixture proportions. The programs can generate a list of genotypes that can be used to help interpret the mixture. Previous Studies Focusing on the Recovery of DNA from Deflagrated Pipe Bombs Previous research at Michigan State University has focused on the recovery of DNA from deflagrated pipe bombs. The experimental design in all of these studies involved the decontamination of pipe bombs with bleach and/or ultraviolet (UV) irradiation to rid them of preexisting DNA prior to volunteer handling. Esslinger et al. (2004) investigated typing recovered DNA from exploded pipe bombs using a standard STR kit (Profiler). Pipe bombs were handled by volunteers, deflagrated, and the remaining fragments collected and swabbed. Out of 20 pipe bombs, four yielded partial profiles that were consistent with the donor’s known profile. Another eight profiles were consistent with the donor’s known profile; however, they were below the reportable threshold. Foran et al. (2009) next investigated the use of mitochondrial DNA (mtDNA) to identify the assembler of a pipe bomb. Results showed that approximately 50% of assemblers were correctly identified. Another 19% were assigned to a group of three individuals that shared the same haplotype, and only one bomb was incorrectly assigned. These results were further supported by mtDNA analysis done by Kremer (2008), where approximately 60% of bomb 7 samples were correctly assigned to a single donor. However, mtDNA is not individualizing; therefore, Kremer (2008) also examined the use of two sets of miniSTRs, miniSGM (http://www.cstl.nist.gov/div831/strbase /miniSTR.htm) and NC01 (Coble and Butler, 2005). Approximately 50% of bomb samples were correctly assigned to a single donor. When combining mtDNA and miniSTR analysis, over 70% of the bombs were correctly assigned to a single individual or a set of two individuals. Once it was established that it was possible to recover nuclear DNA, mtDNA, or both from pipe bombs, subsequent research at Michigan State University focused on optimizing the recovery and analysis procedures. Using miniSTRs, Gomez (2009) compared two recovery methods, the double swab technique (Sweet et al., 1997) and a soaking technique, to find the most efficient way to recover DNA from bomb fragments. For the soaking technique, bomb fragments were placed in a bag with 20 ml of digestion buffer and agitated. The soaking technique was shown to be quicker, but was less effective for DNA retrieval. Soaking made it hard to remove debris from the extract, which co-extracted with the DNA and caused PCR inhibition. Despite being more tedious, the double swab technique was more effective as it generated cleaner extracts and greater DNA quantities. Recovering DNA from similar evidence items has also been investigated. Hoffman (2008) conducted a blind study that focused on the recovery of DNA from IED containers. Decontaminated backpacks were distributed to volunteers who used them in everyday activities for 11 days. The backpacks served as containers for pipe bombs, which were then deflagrated. Instead of using a single swab, multiple swabs of different regions of the same backpack were processed as replicates. These replicates were used to develop a consensus profile for each 8 backpack that was then compared to a reference sample. In the study, all consensus profiles matched the reference profiles, with the exception of a single ambiguous locus. Goals of the Current Study Past IED studies pertaining to the recovery of DNA from deflagrated pipe bombs have all dealt with scenarios where a single individual was handling a pipe bomb/container on which any preexisting DNA had been destroyed using bleach and/or ultraviolet radiation (Esslinger et al., 2004; Hoffmann, 2008; Kremer, 2008; Foran et al., 2009; Gomez, 2009). While controlled conditions are necessary for understanding how DNA is deposited on an IED component and how best to recover it post-deflagration, in actuality the components of a pipe bomb are likely to be handled by multiple individuals, including the manufacturers, shelf stockers at a store, customers, and various other people, before the assembler takes possession of them. This poses the problem of obtaining genetic results from individual(s) other than the handler, potentially resulting in irresolvable mixtures or even erroneous identifications. The goal of the current study was to determine the effects of multiple handlers on the development of DNA profiles for the identification of assemblers of deflagrated pipe bombs. Volunteers were asked to mock assemble pipe bombs that had not been previously decontaminated. Bombs were deflagrated in a controlled environment and the fragments collected and swabbed. DNA was extracted, quantified, amplified, and typed. Consensus profiles were developed blindly for each handler and then compared to reference buccal swabs. 9 MATERIALS AND METHODS Preliminary Analysis of the Effects of Multiple Handlers Using End Caps and Pipe Bombs A preliminary set of experiments was conducted to optimize the design of this study. end caps and pipes were used to investigate isolation and genotyping of touch DNA from multiple handlers, to determine if the last handler (i.e., the perpetrator) could be identified. In these experiments the pipe bombs were decontaminated before handling and were not deflagrated prior to analysis. Each of four end caps was handled by five volunteers, and each of three pipe bombs was handled by three volunteers. Three swabs were used to recover DNA from each end cap, and six swabs were used to recover DNA from each bomb: two per end cap and two per pipe. The double swab technique (Sweet et al., 1997) was used in all instances. DNA was isolated from swabs using a QIAamp DNA Investigator Kit (Qiagen, Valencia, CA) following the manufacturer’s protocol, with the exception of swabs being moistened with 200 µl of Buffer ATL and equal volumes (350 µl) of Buffer AL and Buffer ATL being added after the swabs were combined. DNA was quantified using a Quantifiler TM Human DNA Quantification Kit (Applied Biosystems, Foster City, CA). FAM dye (Bio-Rad, Hercules, CA) and VIC dye (Applied TM Biosystems) were calibrated following the iQ 5 manufacturer’s instructions. Calibrations were performed using 15 µl reactions in 96 well, ThermoGrid® PCR plates (Denville Scientific Inc., Metuchen, NJ) sealed with Microseal® ‘B’ film (Bio-Rad). Quantifiler TM standards were serially diluted as per manual instructions and ranged from 50 ng/µl to 0.023 ng/µl. PCR 10 reactions contained 15 µl volumes made up of 6.3 µl of primer mix, 7.5 µl of reaction mix, and o 1.2 µl of DNA. Thermal cycling parameters consisted of an incubation step at 95 C for 10 o minutes, followed by a denaturing step at 95 C for 15 seconds and an annealing/extension step at o 60 C for 1 minute, carried out for 40 cycles. Real-time data were analyzed using Bio-Rad iQ TM 5 software version 2.1. TM DNA was amplified using an AmpFlSTR® MiniFiler PCR Amplification Kit (Applied Biosystems). Ten microliter reactions consisted of 4 µl of MiniFiler TM µl of MiniFiler TM Master Mix, 1 Primer Mix, and 5µl of a combination of the extracted DNA and low (10mM Tris, 0.1mM EDTA, pH 8.0). The target amount of DNA in each reaction was 0.5 ng. If the DNA quantity data indicated that 5 µl or more of the extract was needed to obtain 0.5 ng, then 5 µl was added. Positive controls contained 4 μl of control 007 DNA (0.1 ng/μl) and 1 μl of low TE. Negative controls contained 5 μl of low TE. Thermal cycling parameters included: an 11 o o minute incubation step at 95 C, followed by 30 cycles of denaturation for 20 seconds at 94 C, o o annealing for 2 minutes at 59 C, and extension for 1 minute at 72 C, followed by a final hold for o 45 minutes at 60 C. The amplified samples were prepared for analysis on an Applied Biosystems 310 Genetic Analyzer. Each reaction consisted of 0.5 µl of GeneScan TM 500 LIZ® Size Standard (Applied Biosystems), 24.5 µl of formamide, and 1 µl of amplified sample. Allelic ladders were made using the same reaction volumes, with the exception that 1.5 µl of the ladder was added. Electrophoresis parameters included: 5 second injection at 15 kV, 30 minute run time at 15 kV, 11 o and a running temperature of 60 C. The data were analyzed with GeneMapperID® Software version 3.2.1 (Applied Biosystems), with minimum calling threshold of 50 relative fluorescence units (RFUs). Consensus profiles were developed by identifying alleles that were consistent among the electropherograms from a single bomb or end cap. These consensus profiles were then compared to those of the last handler to check for concordance. Pipe Bomb Preparation and Handling Following initial studies, thirty-six, 1 foot long by 1 inch in diameter galvanized steel pipes were purchased from a local hardware store. Each was cut from a longer 10-foot pipe and threaded at the store. Bombs consisted of one piece of pipe and two end caps (Figure 1), one of which had a ¼ inch hole drilled in the center for fuse placement. Each pipe bomb was separately placed in a paper bag until volunteer handling. Bags were labeled 1 – 17, 19 – 35, RB1, and RB2, the latter two of which were not handled by volunteers and therefore served as substrate controls. Volunteers were asked to select a letter and a number prior to handling, both of which were unknown to the investigators. The number corresponded to the pipe bomb and the letter corresponded to a reference buccal swab that each participant gave. Prior to handling, volunteers were given no instructions pertaining to hand washing or use of cosmetic products. Subjects mock assembled the pipe bomb for approximately 1 minute and securely fastened the end cap without the hole to the pipe. The end cap with the hole was left unfastened from the pipe. Bombs were placed back in their original bags until deflagration. The use of human subjects followed guidelines established by the University Committee on Research Involving Human Subjects (IRB # 07-577). 12 Figure 1. Steel pipe bomb. Steel pipe with attached end caps. For interpretation of the references to color in this and all other figures, the reader is referred to the electronic version of this thesis. Pipe Bomb Deflagration and Collection Pipe bombs were transported to the Operating Engineers Local 324 (Howell, MI), and deflagrations were performed within a cement structure (Figure 2) that was assembled at the facility. Immediately prior to deflagration members of the Michigan State Police Bomb Squad filled the pipes with 1.5 ounces of Green Dot Smokeless Shotshell Powder (Alliant Powder Co., Radford, VA). The end cap containing the hole for the fuse was fastened onto the corresponding pipe. The pipe bomb was placed in a steel crate (Figure 3) and the fuse was lit. After each deflagration, fragments within the crate were collected as well as any fragments in the surrounding area within the cement structure. Collected fragments were placed into their original bags. The bottom of the crate was vacuumed to remove remaining material and prevent cross-contamination between each deflagration. All investigators directly involved in the deflagration process wore disposable gloves, facemasks, and sleeves. 13 Figure 2. Cement structure. The cement structure was made by employees at the Operating Engineers Local 324. It was used to retain any bomb fragments that escaped from the metal crate. Figure 3. Metal crate used to retain pipe bomb fragments. Photo taken by Stephen Gicale The metal crate was supplied by the Michigan State Police Bomb Squad. It was used to limit the dispersal of pipe bomb fragments. The crate was made of steel and ventilation holes were cut in the sides to relieve pressure from the blast. DNA Recovery and Extraction from Deflagrated Pipe Bombs Prior to extraction all labware such as pipettes, tips, and tubes were autoclaved and UV irradiated in a Spectrolinker XL-1500 UV Crosslinker (Spectronics Corporation, Westbury, NY) 2 for 5 minutes per side (approximately 7.5 J/cm ). Pipe bombs were individually processed in a 14 laminar flow hood decontaminated with 10% bleach. Bomb fragments were separated and swabs were taken of the end caps and the pipe (Figure 4). Six swabs were used to recover DNA from each bomb: two per end cap and two per pipe. The double swab technique was used in which a swab moistened with 200 µl of filter-sterilized 5% SDS was passed over the targeted fragments, followed by a dry swab used on the same fragments. Swabs were labeled with the pipe bomb number and an identifier for the location that was swabbed (Table 1). Both swabs (wet and dry) were placed into the same 1.5 ml microfuge tube along with 500 µl of digestion buffer (20mM Tris, 50mM EDTA, 0.1% SDS, pH 7.5) and 4 µl of proteinase K (20 mg/ml). The tubes were vortexed for 10 seconds and incubated overnight at 56ºC. Figure 4. Deflagrated pipe bomb separated into six sections. Representative fragments from each section of the bomb were grouped together and swabbed separately. 15 Table 1. An example of swab labeling using pipe bomb #27. Location on Bomb #27 Label 1/2 of first end cap 27E1 Other 1/2 of first end cap 27E2 1/2 of second end cap 27E3 Other 1/2 of second end cap 27E4 1/2 of pipe 27P1 Other 1/2 of pipe 27P2 Swabs were labeled with the pipe bomb number followed by the identifier for the location. Figure 4 illustrates the swab locations. ‘First’ indicates the first end cap that was swabbed from that pipe bomb. Both swabs were transferred to a spin basket that had been inserted into a fresh 2 ml SpinEze® (Fitzco Inc., Spring Park, MN) collection tube. The swabs were centrifuged for 1 minute at 1,000 x g. The spin baskets and swabs were discarded and the remaining liquid was pipetted back into the original tube. An equal volume of phenol (500 µl) was added to the extracts, which were vortexed and centrifuged for 5 minutes at 21,000 x g. Aqueous layers were transferred to new 1.5 ml tubes and an equal volume of chloroform was added. Samples were vortexed and centrifuged for 5 minutes at 21,000 x g. The aqueous layers were transferred to Amicon® Ultra spin columns 30K (Millipore Corporation, Billerica, MA) followed by a 10 minute centrifugation at 14,000 x g. The extracts were washed twice with 300 µl of low TE and centrifuged for 10 minutes at 14,000 x g. The columns were inverted into new 1.5 ml microfuge tubes and centrifuged for 2 minutes at 1,000 x g. The extracts were further purified using a OneStep PCR Inhibitor Removal Kit (Zymo Research, Foster City, CA). Each extract was 16 transferred to a Zymo-Spin TM IV-HRC column in a clean 1.5 ml microfuge tube and centrifuged for 1 minute at 8,000 x g. Reagent blanks were processed following the same protocol. These blanks contained swabs although in this case, the swabs were UV irradiated and were not used to o swab any bomb components. All DNA samples were stored at -20 C until further analysis. DNA Quantification, Amplification, and STR Analysis DNA was quantified and amplified using the procedures described above. The amplified samples were prepared for analysis on an Applied Biosystems® 3500 Genetic Analyzer. A master mix was made consisting of 10 µl of deionized formamide and 0.3 µl GeneScan TM 500 LIZ® Size Standard (Applied Biosystems) per sample. Nine microliters were aliquotted into each well of a 96 well ABI Prism sample plate (Applied Biosystems) with 1 µl of amplified product or allelic ladder. Electrophoresis settings consisted of the Fragment Analysis 50_POP7_1 instrument protocol, the Fragment_Analysis_Minifiler_Protocol for size calling, and TM a 50 cm capillary. Runs were conducted using POP-7 (performance optimized polymer 7; Applied Biosystems). The following parameters were used: 5 second injection at 1.6 kV, 30 minute run time at 15 kV, and a running temperature of 60°C. Data were analyzed with GeneMapper® Software version 4.1 (Applied Biosystems). The analysis method was Minifiler_GM 4.0, the panel was Minifiler GS500_v1, and the size standard setting was LIZ (100 – 500). The minimum calling threshold was set at 150 RFUs. 17 Processing of Reference Samples Reference buccal swabs from each volunteer were extracted and quantified as described above, and then diluted to the target amount of 0.5 ng. Amplification was carried out following the protocol described above, with 4 µl of low TE and 1 µl of the diluted DNA being added to TM the MiniFiler reaction. Electrophoresis was performed on an Applied Biosystems® 3500 Genetic Analyzer following the protocol above. Development of Consensus Profiles Electropherograms from each swab were manually reviewed, and each locus was examined with specific guidelines for consensus profile development (Buszek and Thomasma, personal communication). First, the peak (allele) with the highest RFU value at a locus was labeled the “base peak.” Next, the peak height ratio (PHR), relative to the base peak, was calculated for all other peaks at each locus. Each allele was then placed in a category based on PHR: Major, if the PHR was ≥ 75% of the base peak, Minor 1 if the PHR was 33 – 66% of the base peak, Minor 2 if the PHR was ≤ 25% of the base peak. For peaks that did not fall directly into a category, their closest existing neighbor was determined by evaluating how close their PHRs were to the major or minor categories. Peaks were placed in the same category as the allele that had the closest PHR. For example, if a locus had major alleles with PHRs of 100% and 89% and had a Minor 1 allele with a PHR of 66%, a peak with a PHR of 71% would be placed in the Minor 1 category because this ratio is closer to 66% than 89%. Once all swabs from the same pipe bomb were analyzed a consensus profile was created following guidelines: Alleles were assigned a score: Major allele = 1, Minor 1 = 0.5, Minor 2 = 0. The scores were totaled from each of the electropherograms where an allele was present, and 18 placed in the consensus profile when equal to or greater than 3. If more than two alleles had a score of 3 or higher, all alleles were placed in the consensus profile. Similarly, in instances where no alleles scored 3, no allele call was made. Finally consensus profiles for each pipe bomb were compared to the reference buccal swabs from the corresponding handler to check for concordance. Once the handler profiles had been determined, each consensus profile was analyzed, and every locus was placed in one of five categories, as follows (Hoffmann, 2008): A. Only the handler’s alleles were present. B. The handler’s alleles were present, but so were others. C. One of the handler’s alleles was present. This involved instances where there may have only been one allele called, or there may have been multiple alleles called, but only one matched the handler’s profile. D. Alleles present, but none of them matched the handler’s profile. E. No alleles were called at the locus. The number of loci from each consensus profile that contained the handler’s alleles was totaled. Containing the handler’s alleles meant that the locus was in category A or B described above. Consensus profiles were then classified as producing all 9, 8, 7, 6, 5, or less than 5 loci with the handler’s alleles (Hoffmann, 2008). Also, electropherograms with two or more loci containing three or more allele calls were designated as mixtures. Statistical Analyses Mean DNA quantities and standard deviations were calculated using Microsoft® Excel 2007 (Microsoft Corporation, Redmond, WA). DNA concentrations that were over three standard deviations away from their respective means were removed. Statistical tests were done using SigmaPlot® version 12 (Systat Software Inc., Chicago, IL). Normality was tested using the Shapiro-Wilk Test. Due to the non-normal distribution of the data, the Mann-Whitney Rank 19 Sum Test was used to compare differences in DNA quantities between end caps and pipes and between consensus profile categories. The Mann-Whitney Rank Sum Test is a non-parametric test that accounts for non-normal data and outliers by looking at ranks and medians (Bachman and Paternoster, 1997; SigmaPlot, 2011). Due to high standard deviation values, the 0 ng/µl quantities were removed from the data and these statistical tests were re-run. P-values < 0.05 were considered statistically significant. 20 RESULTS Preliminary Analysis of the Effects of Multiple Handlers using End Caps and Pipe Bombs Assessment of loci within consensus profiles from each end cap showed that 6 of 36 (17%) contained all of the last handler’s alleles (Table 2). Similarly, 17 of 27 (63%) loci within the consensus profiles from each pipe bomb contained all of the last handler’s alleles (Table 2). Twenty-two of 63 (35%) alleles from each end cap were consistent with those of the last handler, while 38 of 50 (76%) alleles from each pipe bomb were consistent with the last handler’s (Table 2). During the DNA isolation step of the current study, the filter columns from the QIAamp DNA Investigator Kit became clogged with dirt and debris from the deflagration process, not allowing the DNA to pass through. Therefore, an organic extraction was used to isolate DNA from the swabs. 21 Table 2. Consensus profiles compared to last handler profiles in preliminary studies. End Cap 1 Locus D13 D7 AMEL D2S D21 D16 D18 CSF FGA Last Handler -11 -11,12 -XY 23,24 23,24 -31,32.2 -9,12 -18,19 10,13 10,13 -22 Pipe Bomb 1 Con End Cap 2 Last Handler 11 11,12 11 10,11 -XX 19 19,24 28 28,33 -11,12 -14,20 12 11,12 25 23,25 Pipe Bomb 2 Con End Cap 3 Last Handler -11 -9,10 -XX 19 18,19 -28,31 -13 -16 -9,10 -20,24 Pipe Bomb 3 Con End Cap 4 Con 11,12 9 XX 17,25 -9,11 18 10 18 Last Handler 11,12 9,10 XX 17,25 27,30 9,11 17,18 10 17,18 Last Last Last Con Con Con Locus Handler Handler Handler 11,12 11,12 11,12 11,12 12 12 D13 10,11 10,11 11 9,10 10,11 10,11 D7 XX XX XX XX XY XY AMEL 19,24 19,24 24,25 17,25 23,25 23,25 D2S 28,33 28,33 -27,30 31,32.2 31,32.2 D21 11,12 11,12 11 9,11 11 9,11 D16 14,18 14,20 15,16 17,18 14.2,16 16,18 D18 11,12 11,12 10,11 10 11,12 11,12 CSF 23,25 23,25 17,25 17,18 23,25 23,25 FGA Consensus profiles (Con) were compared to profiles from the last handler of each end cap and pipe bomb. Gray boxes indicate loci that contained all of the last handler’s alleles. (--) indicates that there were no callable alleles at that locus. Pipe Bomb Deflagration Over time, a hole developed in the front of the steel crate that was used for deflagrations. This allowed some of the pipe bomb fragments to escape from the crate. Although the cement structure was designed to retain these fragments, occasionally fragments were propelled outside of the cement structure making it difficult to find and collect them. Pipe bomb 10 was missing an entire end cap and three other bombs (21, 24, and 25) were missing half of an end cap 22 (Appendix A). Despite the missing pieces, these four bombs were processed as normal. The level of fragmentation varied among pipe bombs (Figure 5). Generally, end caps fragmented into two large pieces and several small pieces, which came from the top, flat portions of the end caps. The majority of pipes remained as one large piece with some slight fraying on the ends. Figure 5. Pipe bomb fragmentation. Fragmentation varied among all pipe bombs. Some pipe bombs remained fairly intact (A), while others were fragmented into several pieces (B). DNA Recovery and Extraction from Deflagrated Pipe Bombs Taking into account the pipe bombs that had missing pieces, 199 swabs were obtained from the 34 handled pipe bombs. During DNA collection, the swabs turned a blackish color as they became soiled with residue from the explosives and flakes of the galvanizing layer from the pipes and end caps. Most of the debris settled to the bottom of the microfuge tube during the 56ºC digestion step and was not carried through the extraction. When phenol was added a dark purple color occurred in the aqueous layer that remained until the extracts were washed on Amicon® columns. Occasionally, extracts still had a slight yellow hue, which was eliminated during purification with a OneStep TM PCR Inhibitor Removal Kit. 23 DNA Quantities Recovered from Deflagrated Pipe Bombs DNA concentrations from handled pipe bombs ranged from 0 (14 extracts) to 2.59 ng/µl, with an average of 1.18E-01 ± 3.01E-01 ng/µl of DNA recovered (Appendix A). For 152 of 199 (77%) extracts, the DNA concentrations fell below 100 pg/µl, thus 23% of extracts met the TM minimum target quantity of 0.5 ng of DNA for input into the Minifiler reaction. The average amount of DNA recovered from the end caps was 1.16E-01 ± 3.48E-01 ng/µl while the average for the pipes was 9.68E-02 ± 1.25E-01 ng/µl (Table 3). A Mann-Whitney Rank Sum Test gave a p-value of 0.353 (Table 3), indicating that there was not a significant difference between the amount of DNA recovered from end caps and pipes. An average of 3.30E-02 ± 5.95E-02 ng/µl of DNA was recovered from the control bombs. All procedural reagent blanks had DNA concentrations of 0 ng/µl, except for one (8.52E-03 ng/µl). The large DNA concentrations for swabs 28P2 and RB2P1 were repeatable; however, these quantities were not incorporated into data analyses because they were over three standard deviations above their respective means. Despite the high DNA concentrations for each of these swabs, their corresponding profiles still had loci where no alleles were called (Appendix B). 24 Table 3. Mann-Whitney Rank Sum Test for comparison of DNA recovered from end caps and pipes. End Caps Pipes Average DNA 1.16E-01 9.68E-02 Concentration (ng/µl) Standard 3.48E-01 1.25E-01 Deviation 0.353 P-value The Mann-Whitney Rank Sum Test showed that, although more DNA was recovered from end caps than pipes, the difference was not statistically significant. After removal of the 0 ng/µl quantities, the average amount of DNA recovered from end caps was 1.24E-01 ± 3.59E-01 ng/µl while the average for the pipes was 1.05E-01 ± 1.27E-01 ng/µl. The Mann-Whitney Rank Sum Test still indicated that there was not a significant difference between the amounts of DNA recovered from end caps and pipes (P = 0.266). Analysis of Drop-out Based on Locus Electropherograms corresponding to each swab were analyzed for drop-out (absence) of the handler’s alleles. To distinguish between drop-out and a lack of DNA, an electropherogram was labeled as experiencing drop-out if it had some loci with the handler’s alleles and some without. Given this, 111 of 199 (56%) swabs experienced drop-out. The highest percentage (16%) of drop-out occurred at the D21 and D7 loci (Figure 6). 25 Figure 6. Characterization of drop-out based on locus. TM The loci in the Minifiler kit from smallest to largest are Amelogenin, D13, D16, D7, CSF, D2, D18, D21, and FGA. There was no trend between the degree of drop-out and size of locus. Categorizing Mixtures Based on Number of Loci Containing Extra Alleles There were 137 of 199 (69%) electropherograms labeled as mixtures (Appendix B). Further analysis of each mixture allowed them to be broken down into categories based on how many loci they had with three or more allele calls (Figure 7). Sixty-eight percent of mixture profiles had extra alleles at more than half (five or more) of their loci (Figure 7). 26 Figure 7. Categorization of mixtures based on number of loci with extra alleles. Mixture profiles were broken down into categories based on how many loci had three or more allele calls. Categorizing Loci within the Consensus Profiles Based on the Presence of Handler Alleles Categorizing the 306 loci within the consensus profiles based on the presence of handler alleles showed that 203 (66%) loci contained only the handler’s alleles (Category A), and 263 (86%) contained the handler’s alleles to some extent (Categories A, B, and C) (Figure 8; Appendix B). The handler’s alleles occurred most frequently at the CSF locus. 27 Figure 8. Locus categorizations based on the presence of handler alleles. Loci for each consensus profile were assigned to one of five categories. A: Only the handler’s alleles were present. B: The handler’s alleles were present, but so were others. C: One of the handler’s alleles was present. D: Alleles were present, but none of them matched the handler’s profile. E: No alleles were called at the locus. Categorizing Consensus Profiles Based on Number of Loci Containing Handler’s Alleles Categorizing consensus profiles based on the number of loci that contained the handler’s alleles (Categories A or B) showed that 11 of the 34 (32%) consensus profiles had the handler’s alleles at all 9 loci, and overall, 26 (77%) of consensus profiles had the handler’s alleles at over half (five or more) of their loci (Figure 9). 28 Figure 9. Consensus profile categorizations based on number of loci with handler’s alleles. Consensus profiles were classified as producing all 9, 8, 7, 6, 5, or < 5 loci with the handler’s alleles. Consensus Profile Quality Compared to DNA Recovery The average amount of DNA recovered per bomb for each consensus profile category, are shown in Table 4. The 7 and 8 loci and 5 and 6 loci categories were grouped. Consensus profiles that produced all 9 loci with the handler’s alleles had the highest average DNA concentration (2.27E-01 ± 4.82E-01 ng/µl), while consensus profiles that produced < 5 loci had the lowest average DNA concentration (1.64E-02 ± 1.64E-02 ng/µl). A Mann-Whitney Rank Sum Test indicated a significant difference in the amounts of DNA between various categories (Table 5). 29 Table 4. Comparison of consensus profile quality and average DNA recovery. Consensus Profile Category 9 7 or 8 5 or 6 Average DNA Concentration (ng/µl) 2.27E-01 6.52E-02 7.33E-02 <5 1.64E-02 Standard 4.82E-01 6.57E-02 8.39E-02 1.64E-02 Deviation Consensus profiles were classified as producing all 9, 7 or 8, 5 or 6, or < 5 loci with the handler’s alleles. The DNA quantities for each pipe bomb in one of the four categories were averaged. After removal of the 0 ng/µl concentrations, the standard deviations for each category were still the same magnitude as their corresponding average DNA quantity. A Mann-Whitney Rank Sum Test between each category indicated a significant difference in the amounts of DNA between the same categories as shown in Table 5, as well as between category 9 vs 5 or 6. Table 5. Mann-Whitney Rank Sum Test for comparison of DNA recovered within each consensus profile category. 9 7 or 8 5 or 6 <5 9 Significant 7 or 8 Significant* Not Significant 5 or 6 Significant Significant Significant <5 P-values < 0.05 were considered statistically significant. (*) indicates p-values that changed when 0 ng/µl values were removed. Analysis of Allele Calls in Consensus Profiles that were Inconsistent with the Handler’s Assessment of all consensus profiles showed 19 of 548 (4%) allele calls were inconsistent with the handler’s alleles. Of these, 9 (47%) occurred in stutter positions. There 30 was no correlation between size of the locus and number of inconsistent allele calls. There was also no correlation between the number of inconsistent allele calls and the dye color channels. 31 DISCUSSION The increase in IED usage within the US has made it essential to find and optimize methods for identifying the assembler of the device. Previous studies that attempted to recover and type DNA from IED components and containers were all based on scenarios where a single individual handled a pipe bomb that had been decontaminated with bleach and/or ultraviolet radiation to rid it of any preexisting DNA. These controlled studies provided useful information about DNA deposition and recovery, but did not account for multiple handlers that might result in a mixed DNA profile. The current study was designed to analyze DNA from pipe bomb components that were not decontaminated prior to handling, thus allowing the effects of potential multiple handlers on the development and interpretation of DNA profiles to be assessed. Preliminary results showed that more of the last handler’s alleles were obtained from pipe bombs than end caps, which most likely is due to the pipe bombs having a larger surface area. Although no complete consensus profiles were developed, just over half (53%) of all alleles from the end caps and pipes combined were consistent with those of the last handlers. There were instances where no alleles were called, as well as instances where alleles were called that were inconsistent with the last handler’s. Overall, the preliminary studies showed that there is potential to overcome the possible mixtures and to recover alleles from the last handler. The low levels of DNA encountered in this study can produce profiles affected by a number of stochastic effects such as allelic drop-in/drop-out, or increased stutter peaks (Gill and Buckleton, 2010). These have the potential to complicate analysis and interpretation of profiles, interfering with the correct identification of genotypes. Applied Biosystems (2007) recommends that amplification with the Minifiler TM kit include 0.5 ng of input DNA in order to yield 32 consistent, reliable profiles. A majority (77%) of IED extracts did not meet this optimal quantity, which could directly affect the reliability of the resulting profiles. In addition to being LCN in nature, the recovered DNA may also have been degraded, which can make profile interpretation more difficult. A higher average DNA concentration was recovered from the end caps than the pipes; however, the difference was not statistically significant. The higher average concentration may have led to higher standard deviations being correlated with the end caps than the pipes. This variation may be attributed to the differences in how individuals shed skin cells, which will directly affect the quantity of DNA recovered (Farmen et al. 2008). It is likely that there was more variation in the end caps because more skin cells were deposited there. End caps are more textured, containing ridges and edges, which are likely places for skin cells to be deposited, while pipes have a smooth surface, which might not loosen or retain cells as well. Also, during the assembly of a pipe bomb, more force is generally applied to the end caps to fasten them to the pipe, causing the shedding of skin cells. Finally, the amount of force applied to each end cap likely varied among individuals, in turn causing the DNA concentrations to vary. Allelic drop-out was encountered during the analysis of electropherograms, as over half (56%) of electropherograms experienced drop-out of the handler’s alleles. The drop-out encountered in this study may have been a direct result of the stochastic effects commonly noticed during LCN DNA analysis; however, even if large amounts of DNA were deposited they could be reduced to LCN by degradation. During the explosion of the pipe bombs, DNA was exposed to extremely high temperatures, which cause its degradation (Baneriee and Brown, 2004). Typically, larger loci are more affected by degradation; therefore, more drop-out would be expected at the larger loci (Balding and Buckleton, 2009). Results from this study showed 33 that the highest percentage (16%) of drop-out occurred at the D21 and D7 loci. D21 is the second largest locus in the Minifiler TM kit; however, D7 is the third smallest locus in the kit, but was equal in drop-out percentage to D21. Also, the second highest percentage (14%) of drop-out occurred at the largest locus, FGA, followed by 12% of drop-out at the second smallest locus, D16. These findings show that there was no clear trend between the degree of drop-out and amplicon size. In this study, DNA mixtures were anticipated, as pipe bombs were not exposed to any cleaning procedures. There are various criteria for labeling a profile as a mixture. Some researchers feel that if a single locus has extra peaks then the profile is a mixture, while others feel that a majority of loci within a profile need to have extra peaks (Budowle et al., 2009). If profiles only contain extra peaks at one locus, there is a chance that this is random drop-in. However, if extra peaks are present at a majority of loci, it provides more confidence when designating the profile as a mixture. In the current study, a majority (68%) of all mixed profiles had three or more allele calls at over half (five or more) of their loci. This finding indicates that these profiles were most likely true mixtures, comprised of DNA from multiple contributors. The remaining mixed profiles had extra alleles at two, three, or four loci; therefore, they were most likely mixtures as well. There are a number of factors that could have potentially contributed to the mixtures observed in this study. The recovered DNA could have come from individuals who previously handled the bomb components, such as manufacturers or shelf stockers. However, based on the concept of secondary transfer, it is also possible that the recovered DNA came from individuals who had no contact with the bomb components. Farmen et al. (2008) investigated secondary transfer by setting up various handling scenarios that involved person to person and person to 34 object transfer. Their study demonstrated that transfer of DNA from one individual to another and to a subsequent object was possible. They also show that the individual who last handled an object was not always the one with the most dominant profile. In relation to the present study, this means that the recovered DNA could have come from an individual with whom the handler had direct contact, as with a handshake, or from an individual who touched an object that the handler had contact with, such as a doorknob. Researcher contamination could also have contributed to the mixtures in the current study. Normally, assessment of the control pipe bombs would be a good indicator of contamination. However, the control bombs were not decontaminated and were expected to contain DNA. The control bombs did contain similar alleles and in some instances they were consistent with the researchers or investigators involved with the study. However, the alleles could not be assigned to a single contributor because it is possible that the researchers/investigators had alleles in common with the previous handlers. Since the number and identity of the previous handlers was unknown, it would be difficult to determine exactly who the alleles originated from. As a result, the control bombs in this study were used only to assess the background levels of DNA originating from any previous handlers. Researcher contamination was then investigated through the use of reagent blanks. There was only a single allele call among these blanks, which was not consistent with any of the researchers, indicating that no contamination was being introduced during processing. Despite the fact that consensus profile development was complicated by having LCN, degraded, and mixed DNAs, handler’s alleles were still identified. Categorization of each locus within consensus profiles showed that 66% contained only the handler’s alleles. This means that the consensus profiles containing those loci had some alleles that were entirely consistent with 35 the handler’s and could be used for comparison to the handler profile. A majority (86%) of the time loci contained at least some useful information; that is, only the handler’s alleles were present, handler alleles were present along with others, or one handler allele was present. Of all consensus profiles, 32% had the handler’s alleles at all nine loci, which means that in these instances the handler was identified. A majority (76%) of consensus profiles had the handler’s alleles at over half (five or more) of their loci, making them valuable for searching against a DNA database. Any number of alleles within a profile can be searched; however, the fewer the alleles the less discrimination there will be among individuals. Therefore, consensus profiles containing the handler’s alleles at over half of their loci could provide sufficient information for comparison to a suspect profile. In the consensus profiles, 4% of allele calls were inconsistent with those of the handler. Of these, approximately half occurred in stutter positions. Normally stutter products are easily identified; however, when dealing with LCN DNA the height of stutter peaks is often increased, causing them to be mistaken for an actual allele peak (Budowle et al., 2009). Due to the complications of interpreting stutter in these scenarios, it is also possible that these peaks were true alleles belonging to individuals other than the handler. It is not known to what extent the inconsistent alleles originated from the researchers or from individuals who had previously handled the bomb components because it is possible that the researchers had alleles in common with the previous handlers. There are a number of factors that could have contributed to the variation in consensus profile quality. Research conducted by Farmen et al. (2008) showed that profile quality can depend on an individual’s tendency to shed epithelial cells, often referred to as shedder status. An individual may be a good, medium, or poor shedder, correlating to the quality of the profile 36 that may be recovered from the deposited DNA. Phipps and Petricevic (2007) investigated factors that influence the transfer of DNA to handled objects and the process of ‘shedding’ by having volunteers hold plastic tubes with each hand and participate in a series of hand washing experiments. This study showed that assigning a shedder status may be difficult, as there are many factors affecting the transfer of DNA to objects, such as time since last washing or hand dominance. In the current study, there was a great deal of variation in the DNA quantities recovered from each pipe bomb, which could be directly attributed to the differences in how individuals shed skin cells or the factors that affect shedding. The specific substrate being handled may have also contributed to variation in consensus profile quality. Goray et al. (2010) also investigated factors that affect the transfer of DNA and found that it can be dependent on the type of substrate being handled or the manner of contact between surfaces. For example, in the current study, end caps have a more textured surface than pipes; therefore, greater cell deposition may be expected. Also, more force is generally applied to end caps than pipes, which would create friction and shedding of skin cells. Profiles containing more of the handler’s alleles had significantly higher average DNA concentrations. There were high standard deviations in average DNA concentrations corresponding to each of the consensus profile categories, which again, is most likely due to the variation in how individuals shed skin cells. Despite the trend of more DNA producing better quality profiles, it is interesting to note that of the 14 swabs that had DNA concentrations of 0 ng/µl, 11 still produced allele calls at all or several loci, some of which were consistent with the handler. An explanation for this could be the differences in sensitivity of the fluorescence detectors used for DNA quantification and STR analysis. Therefore, it is possible for DNA to not be detected during quantification, but for alleles to be detected during STR analysis. The 37 dyes used in the Quantifiler TM TM and Minifiler kits were analyzed for any differences that may be causing results at one stage and not the other. The same FAM TM and VIC® dyes are used in each of the kits; therefore, there is no difference between them. The Minifiler TM contains the NED and PET TM TM dyes, whereas the Quantifiler TM kit also kit does not, so the performance of these two dyes could not be compared between the two kits. In contrast, there were also some instances where a swab had a DNA concentration, but TM no alleles were called during STR analysis. In some cases Quantifiler TM the amount of DNA present. The Quantifiler TM the loci amplified in the Minifiler kit targets a 62 bp locus, which is smaller than kit. Smaller loci are less affected by degradation when exposed to extreme conditions; therefore, it is possible for the Quantifiler TM detected, but for the larger Minifiler may be over-estimating TM locus to be loci to not amplify due to degradation. Although using detailed guidelines aided in consensus profile development, the current guidelines are still being investigated. There are gaps between the ranges of PHRs corresponding to each category, and in some instances the PHR for an allele fell outside those ranges. Typically in these instances alleles were placed in the same category as their ‘nearest neighbor,’ or the allele that had the closest PHR; however, sometimes the next closest allele was the base peak. An example from this study was where an allele had a PHR of 31% and its nearest neighbor was the base peak, with a PHR of 100%. This allele was difficult to categorize because the PHR did not seem large enough to place it in the Major category, even though that is the category of the next nearest allele. Individual discretion is used in this situation, which 38 introduces subjectivity in developing consensus profiles. A possible solution would be to extend the PHR ranges so that there were no gaps; however, this could lead to situations where alleles may have similar PHRs, such as 74% and 76%, but are placed into different categories. Scoring of alleles is another aspect of the consensus profiling guidelines that could be further investigated. For a majority of loci placed into categories C, D, and E (one of the handler’s alleles, none of the handler’s alleles, or no callable alleles), the handler’s alleles were in the replicate electropherograms from each swab; however, they did not have a score of three or higher, and therefore were not included in the consensus profile. Also, instances occurred in which a locus had more than two alleles that scored a three or higher, therefore they were all included in the consensus profile. To address these situations the cut-off score for consensus profile placement could be modified, although problems may still be encountered. If the score was lowered, more of the handler’s alleles may be included in the consensus profile; however, this would also allow other inconsistent alleles to be included. In contrast, if the score was increased, the number of allele calls inconsistent with the handler’s may be reduced, but this could also cause some of the handler’s alleles to be excluded from the consensus profile. As was done in this study, the cut-off score might be based on the number of swabs being used, depending on the item being swabbed. For example, items with a larger surface area would require more swabs; therefore, the cut-off score would be increased. Before these consensus profiling guidelines could be implemented, further optimization and validation is needed. Optimization studies could be conducted using known amounts of nondegraded DNA, which would show how the guidelines perform under optimal conditions. For example, the guidelines could be tested on replicate profiles from known amounts of DNA that had been spotted on the same item. Next, tests could be done decreasing the quantity or quality 39 of the DNA to monitor how this would affect the ability to develop profiles using these guidelines. To decrease quality, the DNA could be artificially degraded. The process of analyzing DNA recovered from non-cleaned bomb components shows promise for application within forensic laboratories; however, at the current stage there are still some limitations. The guidelines for processing biological samples can vary among forensic laboratories. Currently, many laboratories will not conduct LCN DNA analysis because the results can be unreliable, which makes testifying in court difficult (Gill and Buckleton, 2010). Also, the guidelines for mixture analysis vary depending on the laboratory; some laboratories may try to interpret the mixture profile, while others may just note that it is a mixture and take no further steps. There is a great deal of research being conducted on techniques to enhance results from LCN DNA or to aid in the process of mixture interpretation (Balding and Buckleton, 2009; Budowle et al., 2009). This study showed that handlers were identified even when dealing with mixed profiles comprised of LCN DNA; therefore, it is possible that in the future forensic laboratories could adopt procedures to routinely swab bomb fragments and utilize variations of these consensus profiling guidelines during profile development. CODIS is a common tool utilized by forensic laboratories. A CODIS profile contains 13 core loci and that can be used for comparison. In the current study, only eight STR loci were targeted for each DNA profile. Future studies could utilize an amplification kit that targets all 13 CODIS loci so that the consensus profiles would have the potential to be uploaded into the database. Searching the consensus profiles within the database could provide useful leads in a case. Although the consensus profile may only be a partial, this could still provide sufficient information to exclude an individual as a suspect. Even in scenarios where the consensus profile may belong to an individual who had contact with the bomb components prior to the assembler 40 that could still help to identify characteristics about the bomb components such as where they were sold or produced, which in turn could link them to a potential suspect. There are also some problems that can arise from searching a consensus profile against a database. For example, if the developed consensus profile was actually a mixture, the profile could randomly match that from an individual already in the database, but who had no involvement with assembling the bomb. Despite the potential problems, it would still be necessary to do a database search to gain any possible information. 41 CONCLUSION This research has shown that it is possible to develop handler profiles from deflagrated pipe bomb fragments, even when the bomb components have not undergone any prior cleaning procedures. As expected, processing bomb fragments that had not been decontaminated did lead to mixed profiles, as over half of the electropherograms were classified as such. Implementing a set of guidelines for consensus profiling was effective in helping to overcome drop-out and mixtures; however, modifications may need to be made to the protocol depending on the particular study parameters. Although a majority of extracts were LCN in nature, informative profiles were still obtained that could be used to help exclude an individual as a suspect. Overall, the results from this research showed that a majority of consensus profiles contained enough alleles to make them useful for searching against the CODIS database, which could provide valuable information. 42 APPENDIX A. DNA QUANTITIES RECOVERED FROM PIPE BOMBS 43 APPENDIX A. DNA Quantities Recovered from Pipe Bombs Extracts were labeled as follows: the first position was the pipe bomb number; “E” denoted an extraction from an end cap; “P” denoted an extraction from the pipe; the last position denoted extraction number for the particular pipe bomb location. All DNA quantities are reported in ng/µl. RB1 and RB2 denoted two pipe bombs that were not handled by volunteers and therefore served as substrate controls (reagent blanks). Bolded numbers highlight extractions that showed no DNA quantities. 44 Table 6. DNA Quantities Recovered from Pipe Bombs. Swab Pipe Bomb 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 E1 3.65E-02 4.16E-02 7.06E-02 1.11E-02 7.99E-02 1.36E-01 1.86E-01 3.20E-02 1.55E-02 2.37E-02 6.41E-02 4.30E-02 8.57E-03 4.99E-02 6.13E-04 1.51E-02 8.43E-03 E2 1.24E-01 7.01E-03 1.97E-02 6.42E-02 1.90E-02 9.65E-02 1.34E-01 6.90E-02 3.12E-02 3.99E-02 4.01E-02 1.25E-02 9.32E-03 8.07E-02 1.08E-02 1.90E-02 1.62E-02 E3 0.00E+00 2.41E-02 1.01E-01 1.06E-01 3.61E-02 1.06E-02 1.67E-01 8.07E-02 9.57E-03 -3.09E-02 7.81E-02 7.43E-03 5.17E-02 1.75E-02 1.08E-02 4.33E-02 45 E4 1.92E-02 1.59E-02 5.09E-02 1.27E-01 7.59E-02 5.83E-02 1.35E-01 6.25E-02 6.79E-02 -4.07E-02 7.42E-02 1.88E-02 8.08E-02 2.89E-03 1.98E-02 3.91E-02 P1 1.51E-03 4.57E-03 3.07E-02 1.11E-01 5.33E-02 7.28E-02 4.80E-01 3.24E-02 2.70E-01 4.30E-03 1.01E-01 9.56E-02 7.32E-02 4.11E-01 2.95E-02 6.88E-02 2.61E-02 P2 1.08E-02 2.58E-02 2.85E-02 2.34E-01 7.22E-02 1.26E-01 4.20E-01 6.73E-02 1.52E-01 1.11E-02 3.45E-02 5.09E-02 2.17E-02 2.30E-01 2.64E-02 2.59E-01 3.52E-02 Table 6 continued 1.92E-02 8.72E-03 4.09E-02 1.19E-02 0.00E+00 3.68E-03 19 5.04E-02 8.35E-02 4.94E-03 1.84E-02 1.48E-01 6.32E-02 20 6.27E-03 4.70E-02 2.18E-02 1.89E-02 1.55E-02 -21 1.83E+00 1.65E+00 2.59E+00 1.91E+00 2.79E-01 2.23E-01 22 4.29E-03 0.00E+00 5.26E-02 4.65E-02 5.40E-02 1.84E-01 23 1.11E-01 5.30E-01 3.12E-01 5.09E-01 9.66E-02 -24 7.08E-03 1.28E-02 9.93E-03 6.20E-03 0.00E+00 -25 0.00E+00 3.35E-02 1.87E-01 2.58E-01 0.00E+00 8.89E-03 26 8.95E-03 0.00E+00 2.58E-02 2.72E-02 0.00E+00 1.59E-02 27 3.93E-03 0.00E+00 1.03E-02 3.78E-03 2.19E-02 1.01E+06* 28 8.88E-02 1.11E-01 1.16E-01 1.11E-01 3.45E-02 6.21E-02 29 4.58E-02 6.37E-02 4.87E-02 2.66E-02 3.34E-03 7.15E-03 30 1.08E-01 2.91E-01 6.37E-02 1.09E-01 3.66E-01 2.44E-01 31 2.38E-02 2.21E-02 2.34E-02 0.00E+00 7.21E-02 4.00E-02 32 5.93E-02 2.76E-02 0.00E+00 0.00E+00 6.98E-03 1.53E-02 33 1.28E-01 2.57E-01 1.39E-01 7.39E-02 5.87E-02 2.44E-01 34 2.98E-03 8.33E-03 1.70E-02 1.05E-02 0.00E+00 0.00E+00 35 0.00E+00 1.53E-01 1.12E-02 3.97E-03 0.00E+00 2.60E-02 RB1 2.29E-03 1.53E-03 0.00E+00 0.00E+00 6.55E+06* 0.00E+00 RB2 *—DNA quantity was not incorporated into data analyses due to its highly skewed value. (--)— indicates that the given pipe bomb did not have a swab from that location. 46 APPENDIX B. ALLELE CALLS AND LOCI CATEGORIZATIONS 47 APPENDIX B. Allele Calls and Loci Categorizations The tables show alleles calls made for each pipe bomb in the study. These calls reflect alleles that made it past the initial analysis of each electropherogram and were placed in either the Major or Minor 1 categories (see Materials and Methods). Asterisk (*) symbols denoted electropherograms that were classified as mixtures during initial analysis. (--)―indicates that there were no callable alleles at that locus. Numbers in parentheses denoted minor alleles. “Con” denoted the consensus profile that was formed after the review of all loci. “Indiv” denoted the handler’s genotype. “Loc Cat” denoted one of the five categories describing the results for each locus of the consensus profile: A. Only the handler’s alleles were present. B. The handler’s alleles were present, but so were others. C. One of the handler’s alleles was present. This involved instances where there may have only been one allele called, or there may have been multiple alleles called, but only one matched the handler’s profile. D. Alleles present, but none of them matched the handler’s profile. E. No alleles were called at the locus. Allele calls for the control pipe bombs (RB1 and RB2) are included; however, their data were not incorporated into analyses. 48 Table 7. Allele Calls and Loci Categorizations for Pipe Bomb 1. Pipe Bomb 1 Locus E1* E2 D13 11,(12) (11),12 11 -D7 XX -AMEL 16,(19), (16),24 D2 (24) -D21 (28),29 D16 10,12 D18 16,17 CSF (10),12 19,(21), (23) FGA E3* E4* (11),12 11 XX (11),12 8,11 XX P1* P2* (11),12 11,12 (10),11 11 XX XX 16,(19), (16),(17), 16,24 16,(20), 24 (24),(25) (20),24 28,(31) (28),31 28,(31) 28,(31) (9),10,12, (9),(10), 10,12 (9),10,13 10,(12) (15) (11),12 (14),(15), 16,17 (15),16, 17 (13),16 16,(17) 16,17 (10),12 10,12 10,12 10,12 10,12 19,(21), -19,(23) 19,23 19,23,33.2 (23) 49 Con Indiv Loc Cat 11,12 11,12 11 11 XX XX A A A 16,24 16,24 A 28 28,31 C 10,12 10,12 A 16,17 16,17 A 10,12 10,12 A 19,23 19,23 A Table 8. Allele Calls and Loci Categorizations for Pipe Bomb 2. Pipe Bomb 2 Locus E1 E2 D13 8 -- D7 AMEL 11,12 XX D2 18 Indiv Loc Cat 8,(13) 8,13 8 8,13 C 11,(12) XX 11,12 XX 11 XX 11,12 XX C A 18 18 18 18 A --18 18,(19) 8,(11), (13) 10,11 XX 18,(19), (20) 30 9,(11), (12),(15) 11,(12) 11,12 11,12 11,12 A (15),16, (15),(17), 17,(18), 21 (21) (17),21 17,21 17,21 17,21 A 10,12 10,12 10,12 10,12 A 30,32.2 32.2 D16 11,(12) -- 12 D18 17 -- 10,(11), (10),12 12 24 Con 8,(11), (13) (10),11 XX D21 FGA P2* E4* 28,29, (31), (32.2) CSF P1* E3* -- 10,12 (30),32.2 (30),32.2 30,32.2 30,32.2 10,12 20,23 21,23, (24) 23,(24) 50 23,(24) 23 23,24 A C Table 9. Allele Calls and Loci Categorizations for Pipe Bomb 3. Pipe Bomb 3 Locus E1* E3* E4* P1* P2* Con Indiv 11,12 (11),12 11,(12) 11,12 11,12 11,12 Loc Cat A 8,(10),11 8,(11) 8,11 8,(11) 8,11 8,11 A XX XX XX XX XX XX A 17,26 17 17,(26) 17,(26) 17,26 17,26 A E2* 11,12 11,(12) 8,(10), 8 D7 (11) XX AMEL XX 17,(19), 17,(26) D2 (20),26 D13 D21 D16 D18 CSF FGA 29,(30), (28),29, 29,(32.2), (31),(32.2), 29,(33.2) (33.2) 33.2 33.2 (9),(10), 10,(12) 10,12 (10),12 12 12,(13) 12,(13) 12,(13) 12,13 (10),(11), (11),12 (10),11 (11),12 12 21,(22), 21,22 21,(22) 21,(22) (23) 29,33.2 (29),33.2 29,33.2 29,33.2 A 10,(12) 10,(12) 10,12 10,12 A (12),13 12,13 12,13 12,13 A 11,12 11,12 11,12 11,12 A 21,(22) 21,22 21,22 21,22 A Table 10. Allele Calls and Loci Categorizations for Pipe Bomb 4. Pipe Bomb 4 Locus E1 E2* E3* E4 P1* P2* Con Indiv Loc Cat D13 D7 AMEL 8 11 XX 8,(12) 9,11 XX 8,(12) 9,11 XX 8,12 (9),11 XX 8,(11),12 (9),11 XX 8,12 9,11 XX 8,12 9,11 XX A A A D2 24 8,(11) 9,10,11 XX 17,(20), (21),24 17,24 17,(24) 17,24 17,(24) 17,24 17,24 A D21 -- -- 29 29,31.2 C D16 D18 9,12 -- 9 16,17,18 9,12 17,18 9,12 17,18 A A 11,12 11,12 A 24,25 24,25 A 29,(31.2) 29,(31.2) 9,(12) (17),18 CSF (10),12 (10),(11),12 11,12 FGA 23 21,24, (25) 24,25 9,(12) 17,18 31.2 9,(12) 17,(18) 11,(12) 11,(12) 24,(25) 24,(25) 51 29,30, (31.2) 9,(12) 17,18 (10),(11), 12 (23),24, 25 Table 11. Allele Calls and Loci Categorizations for Pipe Bomb 5. Pipe Bomb 5 Locus E1* E2* E3* E4* P1* P2* Con Indiv D13 11 11,(13) (11),12 (11),13 13 11 13 D7 10,11 9 -- 11 (9),10, (11) X(Y) Loc Cat D (9),10 9 9,10 9,10 A XX A 20,25 B 29,31.2 C 11,13 C 12,17 B AMEL D2 D21 D16 D18 CSF FGA X(Y) XX XX 19,(20), 17,(19), (19),20, (25) 20,25 (25) (28),(30), (28),31 29,31.2 31 9,11,(15) 11,(15) (9),11 (12),(13), (12),(14), 12,17, (18) (15),16 15,(16) 10,12, 10 12 (14) (21),23 22 22,(23) 20,(25) 29,(31) 11,(13) 12,17, (15) (10),12 XX XX XX 19,20, 20 19,20,25 (24),(25) 29,(28), 32.2 29 31.2 (9),11,13 11,(13) 11 (12),(13), 12,17 12,15,17 15,16 (10),12 (20),21, (20),21, (22),(23) 22,(23) 12 10,12 12 B 20,22 22 20,22 C Table 12. Allele Calls and Loci Categorizations for Pipe Bomb 6. Pipe Bomb 6 Locus E1 D13 8,(11) D7 (10),12 AMEL XY D2 18,23 E2 E3* E4 8,(11) (10),12 (X)Y 8,11 (10),12 (X)Y 8,(11) 12 XY 18,23 18,23 D21 30 29,30 D16 D18 CSF FGA 12 14 10,(11) 22,24 12 14 10 22,24 30,33.2 P1* 8,(11) 10,12 (X)Y 18,(19), 18,(23) (23) P2* Con Indiv 8,(11) 10,(12) (X)Y 8,11 10,12 XY 8,11 10,12 XY Loc Cat A A A 18,(23) 18,23 18,23 A 30,(31), (31.2), 33.2 30 (9),11,12 12 14 (13),14 (10),11 10 22 -- 52 30,33.2 30,33.2 30,33.2 A 12 14 10,11 22,(24) 12 14 10,(11) (22),24 A A A A 12 14 10,11 22,24 12 14 10,11 22,24 Table 13. Allele Calls and Loci Categorizations for Pipe Bomb 7. Pipe Bomb 7 Locus E1* D13 10,(12) D7 (10),(11),12 AMEL D2 XX (23),24 E2* 8,(12) 14,(15) 10 21,24.2 E4* 10,12 10,(12) 10,(12) 10 10,12 (10),12 XX XX XX (23),24 (23),24 23,24 D21 (31),31.2,(32) 31.2 D16 D18 CSF FGA E3 31.2 32 P1* P2* Con Indiv 10,12 10,12 10,12 10,12 Loc Cat A 10,12 10,12 10,12 10,12 A XX 23,(24) XX 23,24 XX 23,24 XX 23,24 A A 31.2,32 31.2,(32) 31.2,32 31.2,32 A 8,12 8,12 8,(12) 8,12 8,12 8,12 8,12 14,15 14,15 14,(15) 14,(15) 14,15 14,15 14,15 10 10 10 10 10 10 10 21,(23) 21,24.2 21,24.2 21,(24.2) 21,24.2 21,24.2 21,24.2 A A A A Table 14. Allele Calls and Loci Categorizations for Pipe Bomb 8. Pipe Bomb 8 Locus E1 E2 E3 D13 D7 AMEL D2 8 9,10 XX 19,23 8,11 8 XX 19,(23) 8,(11) 9 XX 19,23 D21 29 (28),29 -- D16 13 11,13 (11),13 D18 (15),(16),18 18 18 CSF FGA 11,12 21 E4* P1 P2* 8,11 11 8,(11) 8,(9) 9 8,9 XX XX XX 19,23 19,(23) 19,23 (28),29, 32 29,32 (32) 11,13 11,13 11,13 18 18 18 Con Indiv Loc Cat 8,11 8,9 XX 19,23 8,11 8,9 XX 19,23 A A A A 29 29,32 C 11,13 11,13 A 18 18 A 11,12 11,12 11,12 9,11 11,12 11,12 11,12 21,24.2 21,(24.2) 21,24.2 21,24.2 21,24.2 21,24.2 21,24.2 53 A A Table 15. Allele Calls and Loci Categorizations for Pipe Bomb 9. Pipe Bomb 9 Locus E1 E2* D13 D7 AMEL D2 ------ D16 D18 CSF FGA ----- P1* P2* 11,(12) 10 XX 16,(21) 11,12 10 XX 16,21 11,12 11,12 10 10 XX XX 16 16,21 Loc Cat A A A C 28,30 28,30 28,30 28,30 A 9,(12) 9,12 9,12 9,12 12,(17) 12,(17) 12 12,17 10,(12) 10,12 10,12 10,12 20,(21) 20,(21) 21 20,21 12 10 XX 16 D21 E3 A C A C E4* 12 11 10 10 XX XX 16,(21) 16,(21) 28,30, --(31) 9,(12) 9,12 9,(12) 12,(17) 12 12,(17) (10),12 10,12 10,12 21 -21 Con Indiv Table 16. Allele Calls and Loci Categorizations for Pipe Bomb 10. Pipe Bomb 10 Locus E1 E2 P1 D13 D7 AMEL 8 -XX 11 --- ---- D2 21,24 20,24 -- D21 D16 D18 CSF FGA 31.2 --(9),11 -9 13,(14) (13),14 -10 10 10 22 --- 54 P2 11 8,11 XX Loc Cat E E E 21,24 E 30,31 11,12 13,14 10 22,24 E E E A E Con Indiv ------(18),19, -(24) 30 -12 -14 -(10),12 10 --- Table 17. Allele Calls and Loci Categorizations for Pipe Bomb 11. Pipe Bomb 11 Locus E1* 11 D13 10,(11) D7 AMEL (X)Y D2 D21 D16 D18 CSF FGA 19,20 E2 E3 E4* P1* 9 -XX -8 XX 9,(11) 8,(9) XX 8,(9) 8,10 XX 18,19,20 (19),20 28,(30) -9,(15) (9),12 (13),15, (13),14 16 10,12 (10),12 21,23 23,24 19 30 9,(12) -9 (10),13, (13),14 (14) 10 10,(12) -- P2* Con Indiv 8,(9) 9 8,9 (8),10 8,10 8,10 XX XX XX (17),19, 19,20 19,20 19,20 20 (30),31.2 30,(31.2) 30 30,31.2 9,12 12 9,12 9,12 24 Loc Cat C A A A C A 13,14 13,14 13,14 13,14 A 10,12 10,12 22,(23), (24) 10,12 10,12 A 23,24 22,24 C 22,23,24 Table 18. Allele Calls and Loci Categorizations for Pipe Bomb 12. Pipe Bomb 12 Locus E1* E2* E3* E4* P1* P2* Con Indiv (9),(11), 9,12 9,12 9,(11),12 9,12 9,12 12 (10),11, (11),12 11,12 (11),12 11,(12) 11,(12) 11,12 11,12 D7 (12) XY XY XY XY XY XY XY XY AMEL (17),18 17,18 17,18 17,18 17,18 17,(18) 17,18 17,18 D2 (28),31, 31,31.2 31,31.2 31.2 31,(31.2) 31,31.2 31,31.2 31,31.2 D21 31.2 12 12 12 12 12 12 12 12 D16 12 12 12 12 12 12 12 12 D18 12 12 12 12 12 12 12 12 CSF D13 FGA (9),12 9,12 19,(20) (19),20 19,20 (19),20 55 19,20 19,20,(21) 19,20 19,20 Loc Cat A A A A A A A A A Table 19. Allele Calls and Loci Categorizations for Pipe Bomb 13. Pipe Bomb 13 Locus E1* E2* E3* E4* P1* (8),(10), 10,11, (13) 10 10,(13) 13 (9),10, 10,11 10,11 (10),11 D7 (11) XX XX XX AMEL XX 20,(24), (17),(18), 20,24 24,(26) D2 (26) 24,26 (28),29, 29,30 29,30 D21 (29),30 30,(31.2) 11,(12), (9),11, (13) (9),11,13 11,(13) D16 (13) (15),18, 18,(20) (12),18, 20 D18 (18),20 (20) 10,(12), 10,(12), 10,(11), 10,(13) CSF 13 (13) (13) 21,23 21,(23) 21,23 FGA (21),23 D13 P2* Con Indiv Loc Cat 10,(11), 10,(13) 10,13 10,13 13 A 10,11 10,(11) 10,11 10,11 A XX XX XX XX A (24),26 24,(26) 24,26 24,26 A 29,(30) 29,30 29,30 29,30 A 11,(13) 11,(13) 11,13 11,13 A (12),18, 18,20 18,20 20 10,(12), 10,13 10,13 10,13 13 21,23 21,(23) 21,23 21,23 18,20 A A A Table 20. Allele Calls and Loci Categorizations for Pipe Bomb 14. Pipe Bomb 14 Locus E1* D13 D7 AMEL D2 D21 D16 D18 CSF FGA 10,13 9,12 XX 17,(24) 29,32.2 12,13 14,(16) 11 20,23 E2* E3* 10,13 10,13 9,12 9,(12) XX XX 17,(24) 17,(24) 29,32.2 29,(32.2) 12,(13) 12,13 14,(16) 14,(16) 11 11 20,(23) 20,23 E4 P1* P2* 10,(13) 10,13 10,(13) 9,(12) 9,12 9,12 XX XX XX 17,(24) 17,24 17,(24) 29 29,32.2 29,32.2 12,13 12,13 12,13 14,16 14,16 (14),16 11 11 11 20,(23) 20,23 20,23 56 Loc Cat 10,13 10,13 A 9,12 9,12 A XX XX A 17,24 17,24 A 29,32.2 29,32.2 A 12,13 12,13 A 14,16 14,16 A 11 11 A 20,23 20,23 A Con Indiv Table 21. Allele Calls and Loci Categorizations for Pipe Bomb 15. Pipe Bomb 15 Locus E1 E2* E3 E4 P1* P2* Con Indiv D13 -- (10),11 11 (10),11 10,11 (8),10,11 D7 -- 12 -- -- 10,(12) 10,(11), 12 -- 10,12 E AMEL -- XX -- XX XX XX XX XX A D2 20 (23),24 D21 -- (19),20, (17),(19),(20), 17,(22), 23, 23,24 23,24 23,24 (23) (22),23,24 24 35 D16 D18 --- CSF FGA 10,11 10,11 Loc Cat A A -- -- 9,11 -18 -(10),11, -10,(12) (12),(15) -22 -- (28),(30),(31), 34.2,35 (33.2), (34.2),35 35 34.2,35 C 9 18 9,(11) (15),18 9,11 (15),18 9 18 9,11 18 C A 11 10,11,(15) (10),11,(15) 10,11 11,15 C -- 21.2,22 21.2,22 22 21.2,22 C Table 22. Allele Calls and Loci Categorizations for Pipe Bomb 16. Pipe Bomb 16 Locus E1 E2 E3* E4* D13 D7 AMEL 12 -XX (11),12 8,10 XX 11,(12) 10,(11) X(Y) 11 11 -- D2 D21 D16 (17),21, (20),21, (24),25 25 28,30 30 (16),(18), 19,21 28,32.2 10,(11) (10),11 13,(14), (15) D18 -- 12,(18) (12),(14),16, (18),(20) CSF 10,12 10 (9),(10),12 FGA -- 21,(24) 21,(23) Loc Cat 11,12 11,12 11,12 11,12 A 8,10 8,10 8,10 8,10 A XY XY XX XY C P1* P2* Con Indiv 19,(20) 21,25 21,25 21,25 21,25 28 30 30 28,30 30 (9),15 10,(11) 10,11 10,11 10,11 15 A C (10),12 10,12 10,(12) 10,12 10,12 A 57 21,24 21,24 12 B 12,18 21,23 12,(18) 12,18 A 21 21,24 C Table 23. Allele Calls and Loci Categorizations for Pipe Bomb 17. Pipe Bomb 17 Locus D13 E1* (10),12 10,12 8,(12) D7 AMEL XX D2 E2* 21 8,(12) XX 21,(26) D21 27,30 27,32.2 D16 11,13 (11),13 D18 13,(14) (13),14 CSF (10),11 10,11 FGA 20,(22) (20),22 E3* E4* (8),10, (12) (8),12 8,9,12 -XX XX X(Y) (20),21, (17),(20), (21),26 26 (21),26 27,30, 27,(28), 27,(32.2) 32.2 (30) 11,13 11,13 11,13 13,14, 13,14 13,14 (16) 10,11 10,(11) 10,(11) 20,(21), 20,22, 20,(21), (22),(23) (23) 22,(23) 10 P2* Con Indiv Loc Cat (10),12 10,12 10,12 A 8,10,12 XX 8,12 XX 8,12 XX A A 20,21,26 21,26 21,26 A P1* 10,12 27,(28), 27,30,32.2 27,32.2 B (30),32.2 11,13 11,13 11,13 A 13,(14) 13,14 13,14 A 10,11 10,11 10,11 A -- 20,22 20,22 A Indiv Table 24. Allele Calls and Loci Categorizations for Pipe Bomb 19. Pipe Bomb 19 Locus D13 D7 AMEL D2 D21 D16 D18 CSF FGA E1 E2* E3* E4 11 11 11 11 10 10,(11) (10),13 -XX XX XX XX 16,(17), 16 16 16,(25) 25 (28),31.2, (29),(31.2), (31.2), -32.2 32.2 32.2 8,(11), 8,(13) (8),13 8 (13) (14),15, 15,17 (15),17 15 (17) 12 12 12 12 19 19 19,23 19,(23) 58 P1* P2* Con 11 10 XX 11 10,13 XX 11 10 XX Loc Cat 11 A 10,13 C XX A 16 16,25 16,(25) 16,(25) C 31.2, (28),31.2, 31.2,32.2 31.2,32.2 A (32.2) 32.2 8,(13) 8,13 8,13 13 B (15),17 15,17 15,17 15,17 A 12 19,23 12 (19),23 12 19,23 12 19,23 A A Table 25. Allele Calls and Loci Categorizations for Pipe Bomb 20. Pipe Bomb 20 Locus E1* E2* 9,11 9,11 (8),(11), 8,11,(12) D7 12 XX AMEL XX D13 D2 D21 D16 D18 CSF FGA E3* E4* P1* P2* Con Indiv Loc Cat 9,11 9,11 (9),11 9,11 9,11 A 11 11,(12) (8),11,12 9,11 (8),11, (12) XX 11,12 11,12 A XX XX A 17,23 17,23 A 32.2 32.2 A 13 13 A 15,16 15,16 A 10 10,11 C 20,23 20,23 A -XX XX (17),(19), 17,(23) 17,(23) 17,(23) 17,(16), 23 17,23 23 (31.2), (31.2), 32.2 32.2 32.2 -32.2 32.2 13 (9),13 13 13 13 13 12,(15), (14),15, (12),15, 15,(16) 15,(16) 15,16 (16) (16),(17) 16 10,(11) 10,(11) 10 10 10,(11) 10,(11) 20 20,(23) 20,23 (20),23 20,(22), 23 20,(23) 59 Table 26. Allele Calls and Loci Categorizations for Pipe Bomb 21. Pipe Bomb 21 Locus E1 E3* E4* P1* P2* Con Indiv Loc Cat -- 8,11 8,11 8,(10),11,(13) 10 8,11 8,13 C 10,(11) X(Y) 11,(13) XX 10,(11) X(Y) (17),(19), (24),25 10,(11) X(Y) 10 XX 11 XX D A 17 17 19,25 D 28,(29) 29 29 29,33.2 C D13 -D7 AMEL -- 17 19,(20),(25) 17,(19), (25) D2 D21 32.2 (28),29,31 (28),(29),33.2 D16 -- 9,(10),11, 12,13, (15) 11,13 9,11,12, (13) D18 15 15,(16) 15 15,(18), 19 CSF -- 10,(11),12 (11),12 (10),11, (12) (22),23 (20),21,(22), (23),24 FGA 20 (21),23,(24) 11,12 11,12 11,13 (15),18, 15 15 19 (10),11, 11,12 11,12 12 24 -- 23,24 C A A E Table 27. Allele Calls and Loci Categorizations for Pipe Bomb 22. Pipe Bomb 22 Locus E1* E2* E3* E4* 8,(13) 8,13 8,13 8,13 D13 8,11 8,11 8,11 8,11 D7 XY XY XY AMEL XY 18,26 18,(26) 18,(26) 18,(26) D2 D21 29,30.2 29,30.2 29,30.2 29,30.2 12,13 12,13 12,13 12,(13) D16 12,16 12,16 12,16 12,(16) D18 12,13 12,13 12,13 12,(13) CSF FGA 21,24 21,24 21,24 21,24 60 Loc Cat 8,13 8,13 8,13 8,13 A 8,11 8,11 8,11 8,11 A XY XY XY XY A 18,(26) 18,26 18,26 18,26 A 29,30.2 29,30.2 29,30.2 29,30.2 A 12,13 12,13 12,13 12,13 A 12,16 12,16 12,16 12,16 A 12,13 12,(13) 12,13 12,13 A 21,(24) 21,24 21,24 21,24 A P1* P2* Con Indiv Table 28. Allele Calls and Loci Categorizations for Pipe Bomb 23. Pipe Bomb 23 Locus E1 D13 D7 AMEL 12 -XX (17),20, (24) D2 D21 28 D16 D18 12 12 CSF 10 FGA 20,21 Con Indiv 11,12,13 11,12 11,12 9,10,(11) 9,(10),11 9,11 XX XX XX 17,(18), -- 17,18 17,18 17,18 17,18 19,(25) 28,29, -- 28,(29) 28,(29) 28,29 28,29 (31) -- 11,12 11,12 (9),11,12 11,12 11,12 -12 12 12 12 12 10,(11), -- 10,(12) (10),12 10,(12) 10,12 12 (20),21, -- 21,23 21,(23) 21,23 21,23 (23) 11,12 9,11 XX Loc Cat A A A 17,18 A 28,29 A 11,12 12 A A 10,12 A 21,23 A E2 E3* -- 11,(12) -- 9,(11) -XX E4* P1* P2* 11,12 (9),11 XX Table 29. Allele Calls and Loci Categorizations for Pipe Bomb 24. Pipe Bomb 24 Locus D13 D7 AMEL D2 D21 D16 D18 CSF FGA E1* E3* E4* P1* P2* Con Indiv 9,(11) 9,11 9,11 9,11 9,11 9,11 9,11 9,10 9,(10) 9,10 9,10 9,10 9,10 9,10 XY XY XY XY XY XY XY 24,(25) 24,25 24,25 24,25 24,(25) 24,25 24,25 29,(33) 29,33 29,(33) 29,33 29,33 29,33 29,33 11,12 11,12 11,12 11,12 11,12 11,12 11,12 14,(18) 14,18 14,18 14,18 14,(18) 14,18 14,18 12,(14) 12,14 12,14 12,14 12,14 12,14 12,14 22.2,24 22.2,24 22.2,24 22.2,24 22.2,24 22.2,24 22.2,24 61 Loc Cat A A A A A A A A A Table 30. Allele Calls and Loci Categorizations for Pipe Bomb 25. Pipe Bomb 25 Locus E1 E3 E4* P1* D13 -- 11 11 11 D7 -- 13 10,11,13 11 AMEL -- XX XX X(Y) D2 20 (17),25 D21 -- D16 9,13 P2* Con Indiv 11 11 11 (10),11, 11,13 11,13 13 XX XX XX Loc Cat A A A (16),17, 17,(18), 17,(24), (19),(20), 17,25 17,25 19,20,25 (25) (24),(25) A 30 (28),30, (31) -- 9,(13) 9 9 (12),(14), (15),17, (18) 10,(11) D18 -- 17 (15),16, (17) CSF -- 10 10,(11) FGA 24 21 21,(22), 20,21,25 (23),(25) 62 (28),30, (31.2) 9,(12), (13) 30 30 A 9 9,13 C 17 17 17 A 10,11, (12) 10 10,11 C 21 21 21,25 C Table 31. Allele Calls and Loci Categorizations for Pipe Bomb 26. Pipe Bomb 26 Locus E1 E2* E3* E4* P1* P2* Con Indiv Loc Cat D13 -- 8,(10),14 8,14 8,(11),14 8,14 8,14 8,14 8,14 A 8 XX 17,19 8 XX 17,(19) 8 XX 17,19 -D7 AMEL -D2 17,19 12 14,15 12 14,15 -XX 17,19 (28),(29), 32,33.2 -14,15 10,12 10,12 10,12 10,12 10,12 10,12 A 20 20 20 20 20 20 A D21 -- 32,33.2 (32),33.2 32,(33.2) D16 D18 --- CSF -- FGA -- 12 14,15 10,(11), 12 20 8 8 8 A XX XX XX A 17,19 17,19 17,19 A (29),32, 32,33.2 32,33.2 A (33.2) 12 12 12 A 14,15 14,15 14,15 A Table 32. Allele Calls and Loci Categorizations for Pipe Bomb 27. Pipe Bomb 27 Locus E1 E2 E3* E4 D13 9,12 -- (8),12 -- D7 AMEL --- --- (10),11 XX --- D2 19,(20) (17),19 23 P1* 8,(11), 12,13 --- P2* C -- 10,11 -- XX E E (17),19, (17),19, 19,20 (23),24, 19 19,23 (24) (25) C -- -- 29 -- D16 -- -- 9,11 -- D18 15 -- 14,15 -- CSF -- -- (10),13 10 FGA -- -- 21,22 -63 10,11 XX 12 Loc Cat 8,12 D21 13 Con Indiv 28,29, 28,29 29 29 (30),(32) -9,(12) -- 9,11 (14),15, (14),15, 15 14,15 (16) (18) 10,(11), 10,12 10 10,13 12 (20),21, (20),21, 21 21,22 (22) (22) A E C C C Table 33. Allele Calls and Loci Categorizations for Pipe Bomb 28. Pipe Bomb 28 Locus E1 E2 E3* E4 D13 11 11 11 -- D7 AMEL --- --- --- D2 19 (19),24 -XX 19,20, (24) D21 -- -- 31 -- D16 -- -- 9 -- D18 15 14 15,(16) -- CSF -- -- 10,(12) -- 11,(12), 13 10,(11) XX 19,(20), (24) 28,29, (31), (31.2) 9,11,12, (15) 15,(16), (18) 10,12 FGA 23 23 23 -- (20),21 19 P1* 64 P2 Con Indiv Loc Cat 11,(12) 11 11,12 C --- --- 9,10 XX E E (17),24 19,24 19,24 A 29 -- 29,32.2 E 9 9 9,13 C 15 15 14,16 D 12 (20),(21), 23 -- 10,11 E 23 23 A Table 34. Allele Calls and Loci Categorizations for Pipe Bomb 29. Pipe Bomb 29 Locus E1 E2* E3* E4* P1* P2* Con Indiv Loc Cat D13 -- 9,10 (9),10, (11) 9,10 (9),10,11 10,11 9,10 9,10 A D7 -- -- (8),10,11 -- 8,10 C AMEL -- -- X(Y) -- XY C (17),25 (17),18, 25 18,25 B 28,30 C 9,12 D 17 B 11,12 B 20,21 A D2 D21 D16 D18 CSF FGA 17,18,25 18,25 28,(29), (30) --- (9),11,12 -(17),18, -17 17 (19) (10),11, 11,12 11,(12) 10,11 12 --20,21 20,21 -- 28 28,(29) (8),10, (8),10,11 10 (11) X(Y) XY XX 16,17, (18),(24), 17,25 17,18,25 25 28,(29), 28,29 28 (30),(32.2) (9),11,(12) 11,12 11 15,17,18, (17),18, 17,18 (19) 19 10,(11), 10,11, 10,11,12 (12) (12) 20,21 20,21 20,21 65 Table 35. Allele Calls and Loci Categorizations for Pipe Bomb 30. Pipe Bomb 30 Locus E1* E2* E3* E4 D13 11 11 11 -- D7 -- -- 12 -- AMEL XX -- XX -- D2 (19),20, (17),23, 23,(24) 23 (24) 28,29,31 (29),30, 30,(31), (31.2) 31.2 -- D16 D18 10,(11), (12) (12),15, 16,(18), (19) P2* Con Indiv Loc Cat A 11,(13) (10),11 11 11 9,(10), 9,(11),12 -9,12 E (11),(12) XX XX XX XX A (17),19, (17),(19),23, 23,24, 23,24 23,24 A 24, (25) (25) 28,(29), 28,(29), 28,30 30,31.2 C (30) 30,31.2 (9),10, 9,10,12 10,12 10,12 A (11),12 -- D21 P1* 10 10,(12) -- 12,18, (19) 12,(18) -- 12,(15), 18,(19) 12,18 12,18 12,18 A 10,12 10,12 A CSF 10,(12) 10,11, (12) 10,12 -- 10,12 10,(11),12 FGA 20,(21), (23) 20 20 -- 20 20 20 20 A Table 36. Allele Calls and Loci Categorizations for Pipe Bomb 31. Pipe Bomb 31 Locus E1* E2* E3 E4 D13 D7 AMEL D2 9,12 --17,(25) 9,12 11 X(Y) 17 ---17 9,11,12 --17,(25) D21 -- (30),31 -- -- D16 D18 CSF FGA -12,13 11,(12) -- 13 12,(13) 11,(12) 20,(21) --11 -- -12,(13) 11,12 -- 66 P1* P2* Loc Cat 9,12 9,12 A 11 11 A XX XY C 17 17,25 C Con Indiv 9,(11),12 9,12 11 11 XY X(Y) 17,25 17,(25) (29),30, 30,(31) -- 30,31 31 13 13 13 13 12,13 12,13 12,13 12,13 11,(12) 11,12 11,12 11,12 20,21 20,21 20 20,21 E A A A C Table 37. Allele Calls and Loci Categorizations for Pipe Bomb 32. Pipe Bomb 32 Locus E1 E2 E3 E4 P1* D13 D7 AMEL D2 D21 D16 D18 CSF FGA ---------- ---18,(19) ---10 -- 13 --18,19 29 -15,(16) 10 23 ---------- 11,13 10,11 X(Y) 18,19 29 11,12 15,(16) 10 23,26 P2* Con Indiv (8),11,13 13 11,13 --- 10,11 --XY 18,(19) 18,19 18,19 29 29 29 --- 11,12 15,(16) 15 15,16 10 10 10 23 23 23,26 Loc Cat C E E A A E C A C Table 38. Allele Calls and Loci Categorizations for Pipe Bomb 33. Pipe Bomb 33 Locus E1 E2 E3 E4 D13 D7 AMEL 11,(13) --19,(20), (23),24 (28),32.2 -14 ---- ---- D21 D16 D18 ---(17),19, 20 ---- CSF -- 11,(12) FGA -- 21,22 D2 P1* P2 11,13 -----(17),23, (19),(20),23, -- (20),25 (24) (25) --32.2 -------14 -(10),11, --11 12 ----- 67 Con Indiv ---- 11,13 8,10 XY Loc Cat E E E -- 23,24 E ---- 31.2,32.2 10,11 14 E E E 11 11,12 C -- 21,22 E Table 39. Allele Calls and Loci Categorizations for Pipe Bomb 34. Pipe Bomb 34 Loc Cat A C A Locus E1 E2* E3 E4* P1* P2* Con Indiv D13 D7 AMEL ---- 8,11 8 XY ---- 8,11 8,9 XY 8,11 8,9 XY 8,11 8 XY D2 -- 24,25 24,(25) (8),11 --(19),24, (25) 24,25 24,25 24,25 24,25 A D21 -- 28,30 -- 28,30, (31) 28,30 28,30 28,30 28,30 A D16 D18 CSF FGA ----- 11,12 13,15 10,11 23,25 --10,11 -- -(13),15 10,11 23 11,12 13,15 10,11 23,25 11,12 13,15 10,11 23,25 11,12 13,15 10,11 23,25 A A A A 8,11 8,9 XY 11,12 13,15 10,11 23,25 Table 40. Allele Calls and Loci Categorizations for Pipe Bomb 35. Pipe Bomb 35 Locus E1 E2 E3 E4 P1* P2 D13 -- -- -- -- 11,12, (13) 11,12 D7 AMEL D2 D21 D16 D18 CSF FGA --19 ------ --------- --------- --19,20 ------ --19,(20) 28,30.2 -15,18 (11),12 20 --(19),20 --15,18 (11),12 -- 68 Con Indiv -- 11,12 -10 -XX 19 19,20 -- 28,30.2 -- 9,11 -- 15,18 -- 11,12 -- 20,23 Loc Cat E E E C E E E E E Table 41. Allele Calls for Pipe Bomb RB1. Locus E1 E2 (11),12, 14 D13 11 D7 10,(11) XY Pipe Bomb RB1 E3 E4 P1 11 11 11,(13) 11,12 -- 11 (8),(10), 11 XX Y XY P2 11,(12), (13) XY 10,11 X(Y) (17),19, 19,(20) 17,25 (19),20 19,20 (17),(19),20 D2 (20),(24) 28,(29), 28,(31) 29,32.2 28,(31) 28,(31) 28,31 D21 (30),31 9,(11), 9,(15) 11,12 9,11 9,(15) 9,(15) (12),(13), D16 (15) 15,16 14 15 15,16 15,(16) 15,16,18 D18 10,(11), 10,12 9,12 10,(12) -10,(12) CSF (12) (20),21, (21),23 25 23 21 21,23 FGA (23) Note: Data from this pipe bomb were not incorporated into analyses. AMEL Table 42. Allele Calls for Pipe Bomb RB2. Locus D13 D7 AMEL E1 (8),11 10 XX D2 19 Pipe Bomb RB2 E2 E3 E4 11 -11 10 --XX --(17),19, 17,19 -(20) P1 ---- P2 ---- (19),20 -- 28,(29), --31 (31) 15 (13),15 -15 D16 15 16,(18) -15 D18 10,12 10,(12) --CSF 21,23 --21 FGA Note: Data from this pipe bomb were not incorporated into analyses. D21 69 -- -- 9 15 12 -- ----- REFERENCES 70 REFERENCES TM Applied Biosystems. 2006. Quantifiler TM Quantifiler human DNA quantification kit and Y human male DNA quantification kit. User guide. Applied Biosystems. 2007. AmpFlSTR® MiniFiler TM PCR amplification kit. User Guide. Bachman R, Paternoster R. 1997. Statistical methods for criminology and criminal justice. McGraw-Hill Inc. New York, NY. Balding DJ, Buckleton J. 2009. Interpreting low template DNA profiles. Forensic Science International: Genetics. 4:1–10. Baneriee M, Brown TA. 2004. Non-random DNA damage resulting from heat treatment: implications for sequence analysis of ancient DNA. Journal of Archaeological Science. 31(1):59–63. Belmont PJJr, Schoenfeld AJ, Goodman G. 2010. Epidemiology of combat wounds in operation Iraqi freedom and operation enduring freedom: orthopaedic burden of disease. Journal of Surgical Orthopaedic Advances. 19(1):2–7. Bill M, Gill P, Curran J, Clayton T, Pinchin R, Healy M, and Buckleton J. 2005. PENDULUMa guideline-based approach to the interpretation of STR mixtures. Forensic Science International. 148(2–3):181–189. Bille TW, Cromartie C, Farr M. 2009. Effects of cyanoacrylate fuming, time after recovery, and location of biological material on the recovery and analysis of DNA from post-blast pipe bomb fragments. Journal of Forensic Sciences. 54(5):1059–67. Budowle B, Onorato AJ, Callaghan TF, Della Manna A, Gross AM, Guerrieri RA, Luttman JC, McClure DL. 2009. Mixture interpretation: defining the relevant features for guidelines for the assessment of mixed DNA profiles in forensic casework. Journal of Forensic Sciences. 54(4):810–821. Buszek A, Thomasma S. 2011. Graduate students. Michigan State University. Butler JM. 2006. Genetics and genomics of core short tandem repeat loci used in human identity testing. Journal of Forensic Sciences. 51(2):253–265. Coble MD, Butler JM. 2005. Characterization of new miniSTR loci to aid analysis of degraded DNA. Journal of Forensic Sciences. 50:43–53. 71 Cowen S, Debenham P, Dixon A, Kutranov S, Thomson J, Way K. 2010. An investigation of the robustness of the consensus method of interpreting low-template DNA profiles. Forensic Science International: Genetics. (In Press). doi:10.1016/j.fsigen.2010.08.010. Daly DJ, Murphy C, McDermott SD. 2011. The transfer of touch DNA from hands to glass, fabric and wood. Forensic Science International: Genetics. (In Press). doi:10.1016/j.fsigen.2010.12.016. Department of Defense (DOD). Antiterrorism 3-07.2. Joint publication, 2006. Website: http://file.wikileaks.info/leak/us-antiterrorism-jp3-07-2-2006.pdf. Accessed June 22, 2011. Esslinger KJ, Siegel JA, Spillane H, Stallworth S. 2004. Using STR analysis to detect human DNA from exploded pipe bomb devices. Journal of Forensic Sciences. 49(3):481–4. Farmen RK, Jaghø R, Cortez P, Frøyland ES. 2008. Assessment of individual shedder status and implication for secondary DNA transfer. Forensic Science International: Genetics Supplement Series. 1:415–417. Federal Bureau of Investigation (FBI). Reports and Publications. Terrorism 2002–2005. Website http://www.fbi.gov/stats-services/publications/terrorism-2002 2005/terror02_05.pdf/view?searchterm=domestic. Accessed March 27, 2011. Federal Bureau of Investigation (FBI). 2007. Handbook of Forensic Services. Website: http://www.fbi.gov/about-us/lab/handbook-of-forensic-services-pdf. Accessed June 16, 2011. Foran DR, Gehring ME, Stallworth SE. 2009. The recovery and analysis of mitochondrial DNA from exploded pipe bombs. Journal of Forensic Sciences. 54(1):90–94. Gibbons AJ, Farrier JN, Key SJ. 2003. The pipe bomb: a modern terrorist weapon. JR Army Med Corps. 149:23–26. Gill P, Buckleton J. 2010. A universal strategy to interpret DNA profiles that does not require a definition of low-copy-number. Forensic Science International: Genetics. 4:221–227. Gill P, Brenner CH, Buckleton JS, Carracedo A, Krawczak M, Mayr WR, Morling N, Prinz M, Schneider PM, Weir BS. 2006. DNA commission of the International Society of Forensic Genetics: recommendations on the interpretation of mixtures. Forensic Science International. 160:90–101. Gomez KM. 2009. Recovery and quantification of nuclear and mitochondrial DNA from improvised explosive devices using comparative soaking and swabbing techniques. Masters Thesis. Michigan State University. 72 Goray M, Mitchell JR, van Oorschot RAH. 2010. Investigation of secondary DNA transfer of skin cells under controlled test conditions. Legal Medicine. 12:117–120. Hoffmann SGP. 2008. Investigative studies into the recovery of DNA from improvised explosive device containers. Masters Thesis. Michigan State University. http://www.cstl.nist.gov/strbase/miniSTR.htm, National Institutes of Standards and Technology. http://www.icasualties.org/OEF/index.aspx, 2009. IED Fatalities in Iraq and Afghanistan. Accessed March 27, 2011. http://www.sigmaplot.com/products/sigmaplot/statistics.php, 2011. SigmaPlot 12. Systat Software Inc. Accessed August 22, 2011. Kremer S. 2008. The use of miniSTRs and mitochondrial DNA to identify handlers of pipe bombs. Masters Thesis. Michigan State University. Owens BD, Kragh JFJr, Wenke JC, Macaitis J, Wade CE, Holcomb JB. 2008. Combat wounds in operation Iraqi Freedom and Operation Enduring freedom. Journal of Trauma: Injury, Infection, and Critical Care. 64:295–299. Phipps M, Petricevic S. 2007. The tendency of individuals to transfer DNA to handled items. Forensic Science International. 168:162–168. Perlin MW, Szabady, B. 2001. Linear mixture analysis: a mathematical approach to resolving mixed DNA samples. Journal of Forensic Sciences. 46(6):1372–1378. Ramasamy A, Harrison SE, Clasper JC, Phil D, Stewart MPM. 2008. Injuries from roadside improvised explosive devices. Journal of Trauma: Injury, Infection, and Critical Care. 65:910–914. Schneider PM, Fimmers R, Keil W, Molsberger G, Patzelt D, Pflug W, Rothämel T, Schmitter H, Schneider H, Brinkmann B. 2009. The German Stain Commission: recommendations for the interpretation of mixed stains. International Journal of Legal Medicine. 123:1–5. Smith A. 2011. Improvised explosive devices in Iraq, 2003-09: a case of operational surprise and institutional response. The Letort Papers. Website: http://www.strategicstudiesinstitute.army.mil/pubs/display.cfm?pubid=1064. Accessed June 16, 2011. Sweet D, Lorente M, Lorente JA, Valenzuela A, Villanueva E. 1997. An improved method to recover saliva from human skin: the double swab technique. Journal of Forensic Sciences. 42:320–2. 73