Alexander F Christensen. Sequence, haplotype, and ancestry. FORENSIC SCI SEM, 2015, 5(1): 1–14. FORENSIC SCI SEM ReceivedA inva rielRavebiAcsleeecd icove enfpodltri enm2d0e 461J a1MMn Auaaprraccrrhhiyl 2222000011115555 Peer Reviewed 2157118X.5.1.R 1 Sequence, Haplotype, and Ancestry: Using the Mitochondrial DNA Hypervariable Region to Predict Forensic “Race” Alexander F Christensen, Ph.D. *, a a Joint POW/MIA Accounting Command—Central Identifi cation Laboratory, 310 Worchester Avenue, Hickam AFB, HI 96853. * Corresponding Author: Alexander F. Christensen, PhD, D-ABFA, is a forensic anthropologist at the Joint POW/MIA Accounting Command Central Identifi cation Laboratory (JPAC/CIL). He earned his BA in archaeological studies and classics at Yale College in 1991 and PhD in anthropology at Vanderbilt University in 1998. Before joining the JPAC/CIL, Dr. Christensen taught at Rutgers University–Camden and Augusta State University. His primary research interest is the structure of prehistoric and historical Mesoamerican populations, and he has used osteological, archaeological, linguistic, and ethnohistoric data to investigate this question. He is a member of the American Academy of Forensic Sciences, the American Anthropological Association, and the American Association of Physical Anthropologists, and has published articles in the Journal of Forensic Sciences, Forensic Science International, Human Biology, World Archaeology, Journal of Archaeological Method and Theory, Journal of Human Evolution, International Journal of Osteoarchaeology, and other professional journals. Dr. Christensen is a diplomate of the American Board of Forensic Anthropology (#79). Introduction et al. 2010; Royal et al. 2010). This such study achieved close to 90% chapter focuses on mtDNA, as it is accuracy in prediction of what was In forensic casework, DNA is the type most frequently sequenced referred to as “coarse ethnic group”— primarily used for the individuation from skeletal remains, due to its high that is, Caucasian, African, Asian, and of evidence, or the identifi cation of copy number and good preservation. Hispanic (Lee, Măndoiu, and Nelson remains. This is particularly true for Because mtDNA is a nonrecombining, 2011). autosomal DNA (auDNA), which can uniparentally inherited genome, This chapter takes a step back be used in a positive identifi cation, but correlating it with individual ancestry from these approaches and breaks the is generally the case for Y-chromosome raises a different, if related, set of problem down into three independent (yDNA) or mitochondrial DNA issues from those posed by autosomal questions that must be addressed in any (mtDNA) as well (Edson and DNA. forensic case: Christensen this volume). While A given mtDNA haplotype can • How accurately can a given variation in yDNA and mtDNA actually be assigned to a particular ancestral sequence be assigned to a only places an individual within one or group using either a phylogenetic specifi c haplogroup? more particular groups of paternally or approach (e.g., Lao et al. 2010) or a • How tightly defi ned is the maternally related individuals, it is used computational approach (e.g., Egeland ancestral population with in conjunction with other circumstantial et al. 2004). The fi rst classifi es each which that haplogroup is evidence (such as a passenger manifest) sequence into a broad haplogroup associated? to determine individual identity. But and follows a simple chart or map • What is the correlation what about when individuation is not to determine with which continental between that ancestral possible, either because references population that haplogroup is population and the are not available for comparison or associated. This approach is often circumstances of the particular because no victim names are known? used in studies of population history forensic case? How well can DNA be used to assign and admixture (e.g., Steffl ova at al. Sometimes, these questions an individual to a population group? 2009). The second is a statistical will allow a given case to be placed Several recent studies have model, independent of phylogeny: within a forensically useful ancestral evaluated the degree to which genetic How similar is this sequence to each category with a fair degree of certainty; data correlates with ancestry and different sequence in the reference sometimes they will not. This chapter therefore its predictive value for populations and to which population will consider the three questions in “race” and/or “ethnicity” (e.g., Lao does it show the greatest affi nity? One order, and then specifi cally examine April 2015 ISSN 2157-118X 1 FORENSIC SCI SEM Alexander F Christensen. Sequence, haplotype, and ancestry. FORENSIC SCI SEM, 2015, 5(1): 1–14. mtDNA haplogroups A, B, C, and including several casualties from the particular mutation may well have D, which are of particular interest Korean War classifi ed as “Mongolian,” occurred multiple times and therefore for the casework of the Joint POW/ whose actual ancestors came from be of limited phylogenetic signifi cance MIA Accounting Command—Central China, Japan, and Korea, and others (Behar et al. 2007). Individuals Identifi cation Laboratory (JPAC-CIL). classifi ed as “Malayan” who appear belonging to different haplogroups and Each of these haplogroups can be found to be of Filipino, Chamorro, or Native exhibiting different polymorphisms in both Native Americans and East Hawaiian ancestry. elsewhere in their mitochondrial Asians, and the distinction between genome may bear the same sequence in these two groups is often of forensic Hypervariable Region Sequence the HVR. Some polymorphisms in the relevance in CIL cases. Finally, two and Phylogeny HVR are more stable and are associated case studies will be considered. with particular haplogroups, but most Illustrative examples are drawn Because mtDNA mutates without haplogroups are defi ned on the basis from CIL casework, which consists recombination, haplotypes can all be of coding region polymorphisms that primarily of skeletal remains believed placed within a single phylogeny based are not sequenced in standard forensic to be associated with US casualties upon their evolutionary relationships casework. In some cases, forensic from World War II, the Korean War, (Figure 15.1). Clades within this individuation may require analysis of and the Vietnam War. For CIL military phylogeny are called haplogroups. such single-nucleotide polymorphisms casework, mtDNA analyses are Early studies of mtDNA variation (SNPs) in the CR to differentiate conducted by the Armed Forces DNA at the population level (e.g., Torroni between otherwise identical HVR Identifi cation Laboratory (AFDIL), and et al. 1993) defi ned haplogroups sequences (e.g., Asari et al. 2007; Just the frequency of each mtDNA sequence based on restriction fragment length et al. 2009). is reported against its casework polymorphisms (RFLPs) and assigned Table 1 lists the basic HVR population database (CPD). The CPD them letter designations. As work motifs associated with each lettered supplements the original Scientifi c progressed, it was discovered that haplogroup. To determine in more Working Group for DNA Methods some of these lettered haplogroups detail what haplogroup a particular reference database of 4,839 individuals nested within others, and an increasing HVR sequence might belong to, several (Monson et al. 2002) with additional use was made of sequence data rather references are available. The current sequences from AFDIL’s own analyses than RFLPs (Macaulay et al. 1999; master phylogeny is available at http:// for a total of 10,428, sorted into Ingman and Gyllensten 2001). In the www.phylotree.org (van Oven and forensically relevant subsamples. current system, the letter designations Kayser 2009; references in this chapter For identifi cation purposes, AFDIL are preserved for historical continuity, are to build 14, dated April 5, 2012) compares evidence sequences to but new clades are designated with a and http://www.mtdnacommunity. confl ict-specifi c databases of family series of numbers and letters within org (Behar et al. 2012). Individual reference samples (FRSs)—that is, the original haplogroups, such as polymorphisms may be searched samples taken from maternal relatives B4a1a1a or D4b1a2a1b (van Oven and against these charts to locate of casualties that provide the mtDNA Kayser 2009). The current phylogeny where a given sequence might fall. sequences that those casualties can be is based upon extensive sequencing of mtDNAmanager (http://mtmanager. expected to have had. entire mtDNA genomes, and its broad yonsei.ac.kr; Lee et al. 2008) provides Throughout these examples, outlines are unlikely to change (Behar a web-based application to compare matches to casualty references will be et al. 2012). sequence data to published sequences listed under the racial classifi cations While haplogroups have been the and predict haplogroup assignment, used in those service members’ original focus of population genetics studies, as well as a reference chart of HVR fi les. Individuals who might now be forensic analyses have individuation mutation motifs for each haplogroup classed as Hispanic, for instance, might as their focus. As a result, forensic (http://mtmanager.yonsei.ac.kr/help/ be listed as Mexican, Puerto Rican, laboratories generally sequence the MutationMotifs.pdf; references in this or White; in the latter case, Hispanic hypervariable region (HVR) and, chapter are to the version dated October surnames are noted. The terms used in occasionally, the broader control 2–3, 2011). Haplogroup assignment the records vary over time and between region (CR). The high mutation rate can be double-checked by comparing individuals, and include some that are across the HVR makes it ideal for the sequence data to the set of complete inappropriate at best to modern ears, individuation, but also means that a mtDNA haplotypes made available by 2 Forensic Science Seminar Volume 5 number 1 Alexander F Christensen. Sequence, haplotype, and ancestry. FORENSIC SCI SEM, 2015, 5(1): 1–14. FORENSIC SCI SEM mtDNA Community (12,813 as of the multiple twigs of the tree, but without the amount of phylogenetic variation. April 6, 2012, release). much phylogenetic signifi cance. As Recent work in South and Southeast As a general rule, sequences a result, throughout this chapter, all Asia, for instance, has defi ned a that exhibit few polymorphisms in sequences are listed with an assumed large number of new haplogroups comparison to the Cambridge reference 73G-263G unless stated otherwise. within M, some (but not all) of which sequence (CRS) will be from European Macrohaplogroup R is marked by have distinctive HVR profi les (see twigs of the phylogeny, in H, V, and the loss of 16223T. While this mutation Chandrasekar et al. 2009; Peng et al. U. This is because the CRS belongs does occur in other branches, the 2010). A large number of sequences to haplogroup H2a2a1. In particular, presence of 16223T is a good indicator from African populations are available any sequence that does not exhibit a that the sequence falls within L, M, or (Behar et al. 2008), but it is likely that 73G almost certainly belongs either N (although Behar et al. 2007 report much diversity remains unsampled to R0 (including HV, H, and V) or that 2.5% of all pre-R genomes exhibit there as well. L0, as those are the primary positions 16223C, while 1.1% of R genomes Throughout the process of on the phylogeny where the original have mutated back to 16223T). Within haplogroup assignment, care should 73A has mutated to G. (The same G– these macrohaplogroups, some of the be taken to avoid overspecifi city, A transition also appears in minor branches, particularly those found in especially if the evidence sequence branches L3h1a1 and C4c2, but they Europeans, East Asians, and Native lacks rare polymorphisms. Sometimes, are very rare.) Further, the absence Americans, are quite well defi ned in the a given HVR sequence might be of 73G combined with the presence phylogeny. Others are still very poorly indicative only of a general area of of 72C places a sequence securely in delineated, as they are found in parts of the tree, or even of multiple, distantly HV0 or its daughter, V. Sequences with the world where the number of samples separated branches. As an example, 263A are more common, occurring in to date is very low in comparison to consider the sequence 16223T-16278T- Table 1. Mitochondrial DNA Haplogroups Haplogroup Defi ning HVR Polymorphismsa Simplifi ed Macroclade Structure 16129A, 16187T, 16189C, 16223T, 16230G, 16311C, 146C, 152C, 195C, L0 247A, 263A 16187T, 16189C, 16223T, 16278A, 16311C, 152C, 182T, 185T, 195C, L1 >L1′2′3′4′5′6 247A 16129A, 16148T, 16166G, 16187T, 16189C, 16223T, 16278A, 16311C, L5 >L2′3′4′5′6 > L1′2′3′4′5′6 182T, 195C, 247A L2 16223T, 16278A, 16390A, 146C, 150T, 152C, 182T >L2′3′4′6 > L2′3′4′5′6 > L′2′3′4′5′6 L6 16048A, 16223T, 16224C, 16278A, 16311C, 146C, 152C, 182T, 185C >L3′4′6 > L2′3′4′6 > L2′3′4′5′6 > L1′2′3′4′5′6 L4 16223T, 16311C, 16362C, 195C >L3′4 > L3′4′6 > L2′3′4′6 > L2′3′4′5′6 > L1′2′3′4′5′6 L3 16223T >L3′4>L3′4′6 > L2′3′4′6 > L2′3′4′5′6 > L1′2′3′4′5′6 M 16223T, 489C (control region) >L3 C 16223T, 16298C, 16327T, 249del, 489C >CZ > M8 > M Z 16185T, 16223T, 16260T, 16298C, 249del, 489C >CZ > M8 > M E 16223T, 16362C, 16390A, 489C >M9 > M G 16223T, 16362C, 489C >M12′G > M Q 16129A, 16223T, 16241G, 16311C, 16362C, 489C >M29′Q > M D 16223T, 16362C, 489C >M80′D > M N 16223T >L3 I 16129A, 16223T, 16391A, 199C, 204C, 250C >N1a′c′d′e′I > N1 > N1′5 > N W 16223T, 16292T, 189G, 195C, 204C, 207A >N2 > N Y 16126C, 16231C >N9 > N A 16223T, 16290T, 16319A, 235G >N O 16213A, 16223T >N S 16223T >N X 16189C, 16223T, 16278T >N R None >N HV 73A >R0 > R V 16298C, 72C, 73A >HV0a > HV0 > HV > R0 > R H 73A >HV > R0 > R J 16069T, 16126C, 295T >JT > R2′JT > R T 16126C, 16294T >JT > R2′JT > R F 16304C, 249del >R9c > R9 > R B 16183C, 16189C >R P None >R U None >R K 16224C, 16311C >U8b > U8 > U2′3′4′7′8′9 > U > R April 2015 ISSN 2157-118X 3 FORENSIC SCI SEM Alexander F Christensen. Sequence, haplotype, and ancestry. FORENSIC SCI SEM, 2015, 5(1): 1–14. L0 L1 L5 L2 L6 Africa L7 L4 L3 M1 D1 Americas D* D4 East Asia D2-3 G M* South Asia M27 M28 Melanesia M29 Q* Q2b Australia M42 M10 M21 Southeast Asia E M9 M8 East Asia Z C* C1 Americas A2 A* N9 East Asia Y N21 O/N12 Australia S N2 W N1 West Eurasia I X N* R5 South Asia U2a-d U2e U8a U8b West Eurasia K U3,4,7,9 U1,5 U6 Africa B2 Americas B4a1a Oceania B4* B* East Asia R* R9 F Southeast Asia P3,4,6 Australia P* Melanesia J T HV1 HV0 West Eurasia HV0a V H R0 200 KYr 150 KYr 100 KYr 50 KYr 0 KYr Fig.1 Simplifi ed phylogeography of human mtDNA. Dates are approximate. An asterisk after haplogroup designations indicates “members of this haplogroup not specifi ed on other branches.” 4 Forensic Science Seminar Volume 5 number 1 Alexander F Christensen. Sequence, haplotype, and ancestry. FORENSIC SCI SEM, 2015, 5(1): 1–14. FORENSIC SCI SEM 16362C (here and elsewhere in sequencing of the full CR would case of A and B, the Americas) and this chapter, C-stretch insertions differentiate D/G and L3b1 sequences. others found in Europe (such as T). are ignored). In CIL casework, this Beyond these broad patterns, sequence has been found in multiple Phylogeny and Ancestry however, there is a fair amount of sets of remains from both North and variation. Within any one designated South Korea, as well as World War Once a haplotype is located haplogroup, smaller branches may have II contexts on Pacifi c islands. The within the phylogeny, what can we much more restricted distributions. mtDNAmanager reference list shows say about its ancestry? On a local Some of these distributions may be this as the profi le for haplogroup D4g1, level, population genetic studies have widely disjunctive, as well, as a result which is consistent with Northeast shown a weak correlation between of historical migrations (Table 2). Thus, Asian ancestry, such as Korean or mtDNA and geography because of while haplogroups L1, L2, and L3 are Japanese. In fact, this sequence is the movement of wives between restricted to sub-Saharan Africans and found in over 2% of modern Japanese communities (e.g., Besaggio et al. their descendant populations, specifi c and almost 1% of Koreans (Sekiguchi 2007). On a broader, continental level, branches within them have been et al. 2008; Jin et al. 2006). Does this it does have a high correlation with documented from Slavic groups and mean that all of those individuals geography, as haplogroups branched likely date to prehistoric small-scale actually belong to haplogroup D4g1? off from each other with the spread population movements (Malyarchuk, Not necessarily. With the addition of of modern humans across the world Derenko, et al. 2008). 16189C, this sequence is also consistent (Fig.1; Maca-Meyer et al. 2001). The with G2a; since 16189 has a fairly high very diverse macrohaplogroup L, aside Ancestry and Forensic Relevance mutation rate, it is possible that some from its descendant clades M and N, haplotypes within this haplogroup is restricted in distribution to African Sequence 1 exhibits an extensive would show the same HVR sequence. populations, which exhibit great series of polymorphisms that allow it Given that D and G are related clades diversity and time depth (Behar et al. to be securely placed in the phylogeny: within macrohaplogroup M and show 2008). Macrohaplogroups M and N 16223T-204C-199C, in that order; similar distributions, it might be safer branched off from L3 around the time place it in the N1a′d′e′I branch, with to state that the sequence probably falls of our ancestors’ departure from Africa 152C-16248T-16355T narrowing it to within D or G, and might specifi cally and subsequently diversifi ed across N1a. N1a1a is marked by 16147A and belong to D4g1. Eurasia. A few branches, such as M1 then 16320T. Finally, 207A and the However, this same sequence and U6, indicate back-migrations into reversal of a previous 16172C mutation is found in 11/10,428 individuals in Africa after this date (Olivieri et al. mark N1a1a2. This subclade has only AFDIL’s CPD: two Koreans, four 2006). Within M, several clades have been reported from Europe and has Turkmen, one Asian American, three received their own letter designations actually been sequenced from Neolithic African Americans, and one Hispanic. (C, D, E, G, Q, and Z), while dozens farmers from Germany (Palanichamy et It also matches family references for of others have simply been numbered. al. 2010). Sequence 2 exhibits the same two casualties from the Korean War: Some of these clades are both very 16223T-204C-199C, followed by the one African American and one Korean distinctive and very restricted in 250C-16391A that marks haplogroup American. Are the African American distribution (e.g., Q in Papua and I (a subclade of N1, and similarly and Hispanic matches due to Asian Australia), while others spread across distributed in Europe). Its placement gene fl ow? Probably not. Instead, it much of Asia and even (in the case within I1a1 is determined by the turns out that the same HVR profi le of C and D) the Americas. Within successive mutations 16311C-16172C- can be found in some haplotypes N, there are two broad clades: R and 203A. In the Pacifi c islands, European within L3b1a (Soares et al. 2012). So N*, with the latter also including the maternal ancestry indicates that the without additional sequencing of other haplogroups A, I, O, S, W, X, and Y. remains are likely to come from a regions of the mtDNA genome, all R is in turn divided into R* (including US or Australian casualty who can we can say about this particular HVR B, F, J, P, and T), R0 (including HV, H, eventually be identifi ed by the CIL. sequence is that it belongs either to and V), and U (including K). The R0 Sequences 3 and 4 both exhibit D/G or to L3b1a. Because one of the and U clades are restricted to western the 16223T-16362C that characterizes mutations that defi ne macrohaplogroup Eurasia, while R* and N* include some haplogroups D, G, and some other, M (including D and G) is 489C, branches restricted to Asia (and in the minor branches of M. G3 is marked April 2015 ISSN 2157-118X 5 FORENSIC SCI SEM Alexander F Christensen. Sequence, haplotype, and ancestry. FORENSIC SCI SEM, 2015, 5(1): 1–14. by 16274A, and G3a2 by 143A-152C- sequence 4 perfectly. Haplogroup G is marks Q1. This is an ancient branch 16189C-16265C, placing sequence 3 generally restricted to Northeast Asia, (perhaps as much as 50,000 years securely. Sequence 4 is more diffi cult, and both sequences have been reported old) only found in the indigenous however, as it also exhibits the 16390A from Japanese individuals (Sekiguchi population of Papua New Guinea and that characterizes haplogroup E. et al. 2008; Nohira, Maruyama, and Island Melanesia (Friedlaender et al. However, no sequence within E in the Minaguchi 2010). On a Pacifi c island, 2005). Such a sequence indicates that reference phylogeny exhibits 16260T, such a sequence is most consistent with the remains are almost certainly of which characterizes G2b2b. By the remains belonging to an Imperial indigenous Melanesian origin. contrast, one of the reference sequences Japanese service member. for G2b2b exhibits 16093C-16189C- Sequence 5 is phylogenetically the How Far Back Is That Asian 16390A as private mutations (that most distinctive of all. Haplogroup Q Ancestry? Distinguishing Asian is, differences present in just one is characterized by 16129A-16223T- and Native American Sequences sequence and therefore not used in the 16241G-16311C, and 16144C-16148T- current phylogeny) and thus matches 16265C-16343G-89C-92A-146C Traditional triracial categorizations in physical anthropology lumped Table 2. Selection of mtDNA Subhaplogroups with Disjunctive Distributions Native American and Asian populations Haplogroup Subhaplogroup N/12813a nb Primary Region together as “Mongoloid” (e.g., Brues L1 200 Africa 1990). In fact, the founding gene pool L1b1a8c 0 Europe of Native Americans is a reduced L2 232 Africa L2a1k 2 Eastern Europe subset of northeast Asian variation, as M 1386 Eurasia the Americas were colonized by small M1 80 North Africa groups crossing the Bering Strait who C 423 Northeast Asia C1b-d 211 Americas were largely isolated over subsequent C4c 4 Americas millennia. This is particularly clear C5c1 11 Eastern Europe D East Asia in the mitochondrial genome. Initial D1 50 Americas studies of mtDNA phylogeography D2a 48 Americas defi ned haplogroups A, B, C, and D D3 7 Americas D4b1a2a1b 4 Eastern Europe as the four founding Native American D4e1c 2 Americas types and recognized that they also D4e4b 2 Eastern Europe D4h3a 45 Americas occurred in Asia (Torroni et al. D5a3a 4 Northeast Europe 1993). As the phylogenetic details A 426 Northeast Asia were refi ned, it became clear that A2 277 Americas X 161 Circumpolar the American branches A2, B2, C1, X2a 24 Americas and D1 were distinct from the Asian X2g 1 Americas branches of the same haplogroups, B 452 East Asia B2 122 Americas and that haplogroup X should also be B4a1a1a 67 Polynesia added as a founder. More recently, full U 1208 Europe U2a,b,c,d 28 South Asia genome sequencing has revealed that U6 119 North Africa, Mediterranean extant Native American mitochondrial Z 62 Northeast Asia, South Asia lineages trace back to at least 15 Z1a1a 22 Scandinavia Table 3. Cases from Pacifi c Islands Individual 1 2 3 4 5 Recovery Papua New Guinea Philippines Philippines Papua New Guinea Papua New Guinea location 16144C, 16148T, 16086C, 16147A, 16172C, 16223T, 16189C, 16223T, 16093C, 16189C, 16188T, 16223T, 16223T, 16248T, 16311C, 16362C, Sequence 16265C, 16274A, 16223T, 16260T, 16241G, 16265C, 16320T, 16355T, 152C, 16391A, 189G, 199C, 16362C, 143A, 152C 16362C, 16390A, 195C 16311C, 16343G, 89C, 199C, 204C, 207A 203A, 204C, 250C 92A, 146C, 208C Haplogroup N1a1a2 I1a1 G3a2 G2b2b Q1 Ancestry European European East Asian East Asian Melanesian N/10428a 0 0 0 0 0 Probable origin United States United States Japanese Japanese Papuan 6 Forensic Science Seminar Volume 5 number 1 Alexander F Christensen. Sequence, haplotype, and ancestry. FORENSIC SCI SEM, 2015, 5(1): 1–14. FORENSIC SCI SEM distinct branches, each with Asian (Derenko et al. 2007), and a similar cousins: A2*, A2a, A2b, B2, C1b, Haplogroup A is a subclade sequence has been reported from Korea C1c, C1d*, C1d1, C4c, D1, D2a, of macrohaplogroup M and is (Jin et al. 2006). It thus appears more D3, D4h3a, X2a, and X2g (Perego et distinguished by the HVR profi le consistent with Japanese ancestry al. 2010). Furthermore, at least one 16223T-16290T-16319A-235G. It has a than American, although the recently additional haplogroup, an otherwise restricted distribution in northern Asia discovered New World branches of undocumented branch of M, has been and is found throughout the Americas haplogroups C and D (Table 2) show found in a 5,000-year-old burial from (Achilli et al. 2008; Derenko et al. that other related but long isolated British Columbia (Malhi et al. 2007), 2007). Within haplogroup A, there are branches may still turn up. indicating that it is possible or even three successive splits marked by HVR Individuals 2, 3, and 4 were likely that other as of yet unknown polymorphisms: 152C differentiates recovered from North Korea. The fi rst lineages are, or were, present in the A3, A4, A7, A9, and A11 from A5, A8, bears a 16187T that is characteristic New World, if at very low frequencies. and A10; 16362C differentiates A4; and of A5a; this sequence is common in Nonetheless, the vast majority of within A4, 16111T-146C-152T-153G both Korean and Japanese populations. Native American sequences belong to differentiates A2. Within A2, A2a and The second and third both exhibit the one of the four originally identifi ed A2b are found in both northeast Asia markers of A2 (except for 152T, which branches. and the Americas, while all the other has mutated back to 152C in multiple In CIL casework, recognizing clades, most of which share also 64T, branches within A2). They cannot be whether a given sequence is of Asian are restricted to the New World. For placed more precisely than that, which or Native American origin is frequently forensic purposes, it is safe to consider is not surprising, as A2 contains a broad probative, as the majority of the remains any haplogroup A sequence exhibiting range of variation, consistent with a received by the CIL come from Asian 16111T and/or 64T to be almost small founding population spreading or Pacifi c contexts. While there are US certainly of Native American ancestry, across a wide area and growing casualties of Asian ancestry missing however distant. exponentially (Kumar et al. 2011). in Korea and Southeast Asia, AFDIL Five cases provide examples from This haplogroup is indicative of Native can compare unknown sequences from haplogroup A (Table 4). Individual American ancestry, which is consistent those areas to their family references. 1 was recovered from the island of with their FRS. Individual 5, recovered A sequence of Asian origin that does Tarawa, where thousands of US and from Vietnam, does not match any not match a known Asian American Japanese service members were buried sequences in the CPD or FRS database. casualty almost certainly represents a in 1943. He exhibits the A4 profi le However, despite the lack of database non-US individual, and the remains without any of the distinctive A2 matches, this sequence can also be are therefore not likely to be identifi ed polymorphisms and also has a 200G, securely placed in haplogroup A2. All by the CIL. Any sequence of Native which marks the A4c′d branch. Since three sequences almost certainly belong American origin found in the region, A4d also has 151T, while A4c has no to US service members. by contrast, has a very high likelihood further HVR mutations, he most likely of representing a US casualty who is falls within the latter clade. While no Haplogroup B (hopefully) identifi able. exact match to this sequence could be located in the literature, A4c is a North Haplogroup B is a subclade of Haplogroup A Asian, and particularly Siberian, clade macrohaplogroup R and is characterized Table 4. Haplogroup A Examples Individual 1 2 3 4 5 Recovery Tarawa North Korea North Korea North Korea Vietnam location 16051G, 16111T, 16093C, 16223T, 16111T, 16223T, 16111T, 16223T, 16223T, 16249C, 16290T, 16319A, 16187T, 16223T, 16290T, 16319A, 16290T, 16319A, Sequence 16290T, 16319A, 16362C, 152C, 200G, 16290T, 16319A, 235G 16362C, 16378T, 64T, 16362C, 146C, 153G, 16362C, 146C, 153G, 235G 146C, 153G 159C, 235G 235G Haplogroup A4c A5a A2 A2 A2 Ancestry Northeast Asian Northeast Asian Native American Native American Native American N/10428a 0 1 Kazakh 0 1 Caucasian 0 1 White (Hispanic 1 White, FRS matchesa None None None surname) 1 American Indian April 2015 ISSN 2157-118X 7 FORENSIC SCI SEM Alexander F Christensen. Sequence, haplotype, and ancestry. FORENSIC SCI SEM, 2015, 5(1): 1–14. in the HVR by (a) the change at 16223 was recovered from North Korea that context, US ancestry is the most to the CRS that typifi es R, and (b) and exhibits the basic B4 profi le plausible explanation. 16183C-16189C. Given that both plus 16390A. Unfortunately, this Individual 4 exhibits the HVR 16183 and 16189 exhibit high mutation polymorphism is not signifi cant motif of B5a, including another rates (the former site is not even within the current B4 phylogeny. The distinctive transversion from C to A at considered in the formal phylogeny), sequence does match four Hispanics 16266. This is a haplogroup that is quite it can be diffi cult to distinguish B if no in the CPD, as well as two casualties common in mainland Southeast Asia other mutations are present. Branches who both appear likely to fall within (Peng et al. 2010). Given that these are found throughout East Asia, the that same modern category. It therefore remains were recovered in Vietnam, Pacifi c, and the Americas (Achilli et appears likely that these remains will be they almost certainly represent an al. 2008; Derenko et al. 2012; Li et al. identifi able as those of a US casualty. indigenous individual. 2007; Melton et al. 1995). Within B, the Similarly, individual 2 has the basic The last three sequences present a primary division is between B4, marked B4 profi le plus 150T. This sequence different conundrum. They exhibit a set by 16217C, and B5, marked by 16140C matches 10 Navajo (out of the 146 in of polymorphisms distinctive to B4a1a: and four other mutations outside the the CPD), but it is certainly possible 16189C-16217C-16261T-146C. This HVR. Within B4, 16261T divides B4a, that it might also match indigenous branch appears to have arisen in early g, h, and i from B4b, d, and e (the last Southeast Asians, given how poorly Austronesian populations spreading a minor branch marked by a slew of that region is represented in databases out across islands of Southeast Asia other polymorphisms as well). A simple and the fact that B4c is present in the (Melton et al. 1995). Individual 5 sequence of 16189C-16217C could region (Peng et al. 2010). Without was recovered from North Korea, belong to either B4b or d. The diffi culty additional information, either from but the same sequence has also been here is that B4b includes haplogroup elsewhere in the genome or from the found in remains recovered from a B2, the haplogroup present in the New archaeological context, these remains World War II context in Papua New World, which is differentiated from B4b cannot be securely assigned to either Guinea. Individuals 6 and 7 come by fi ve mutations outside the HVR. American or Asian ancestry. from Saipan and exhibit the 16247G Therefore, HVR data alone cannot Individual 3, by contrast, presents that marks B4a1a1a. The combination always discriminate between Asian and a set of polymorphisms distinctive to of 1627C-16247G-16261T has been American Bs. Fortunately, B4b1, the B2g1. The transversion of C to G at labeled the “Polynesian motif” (Redd largest clade within B4b, is marked by 114 is particularly signifi cant, as most et al. 1995), and it is one of a handful 16136C, usually with multiple other mutations are transitions (between of lineages carried by the Austronesian polymorphisms as well, which does either C and T or A and G). This is a populations that expanded into allow its separation from B2. haplotype that appears to be of native Polynesia (Kayser 2010). Table 5 provides some examples Mesoamerican origin. The remains At fi rst glance, the population of the diffi culties raised by haplogroup were recovered from a World War II database matches of these three B in evidence cases. Individual 1 aircraft crash site in Germany, and in sequences do not match what might Table 5. Haplogroup B Examples Individual 1 2 3 4 5 6 7 Recovery North Korea Cambodia Germany Vietnam North Korea Saipan Saipan location 16182C, 16183C, 16182C, 16183C, 16182C, 16183C, 16183C, 16189C, 16140C, 16183C, 16182C, 16183C, 16183C, 16189C, 16189C, 16217C, 16189C, 16217C, Sequence 16189C, 16217C, 16217C, 16298C, 16189C, 16266A, 16189C, 16217C, 16217C, 150T 16247G, 16261T, 16247G, 16261T, 16390A 114G, 146C 210G 16261T, 146C 146C 146C, 151T Haplogroup B4/B2 B4/B2 B2g1 B5a B4a1a B4a1a1a B4a1a1a East Asian/Native East Asian/Native Island Southeast Ancestry Native American Southeast Asian Polynesian Polynesian American American Asia 4 Asian, 2 Chinese, 4 African 2 Asians, N/10428 4 Hispanic 10 Navajo 5 Hispanic 0 1 African American, 1 Hispanic, American 1 Hispanic, 1 other 1 other 1 Black, 1 White, 1 White (Hispanic 2 Hawaiian, 1 White (Hispanic 3 Hawaiian, FRS matches surname), None None 2 Filipino 1 Chinese surname) 1 Chinese, 1 Mexican Hawaiian, 1 “Mongolian” 1 Asian American 8 Forensic Science Seminar Volume 5 number 1 Alexander F Christensen. Sequence, haplotype, and ancestry. FORENSIC SCI SEM, 2015, 5(1): 1–14. FORENSIC SCI SEM be expected from their phylogenetic (Allard et al. 2005; Diegoli et al. 2009; as examples (Table 6). position. The “Asian” samples are all Steffl ova et al. 2009). These sequences Individuals 1 and 2 exhibit either Taiwan Chinese, who could be are usually attributed to recent non- sequences that can be securely of indigenous Austronesian ancestry, or African admixtures. However, it is placed within the phylogeny. The Asian Americans, a category that often more likely that they trace back to fi rst has the standard C1 motif plus includes Pacifi c Islanders as well. But Madagascar, one source of the Atlantic 16086C-16189C-16278T-143A, which how are distinctly Pacifi c sequences slave trade (Lee et al. 2009; see together mark C1b4. The second appearing within the African American Razafi ndrazaka et al. 2010 for a recent exhibits the C1 motif minus 16223T and Hispanic populations? Some insight analysis of Malagasy mtDNA). (an example of the parallel occurrence is provided by the list of matching So what is the signifi cance of of the T–C transition at this locus). casualties from within the FRS a B4a1a or B4a1a1a sequence in The 215G matches the profi le expected database, for whom some individual CIL casework? It depends upon the for C1c1. Both of these sequences are biographical data are available. recovery area. For remains recovered present in CPD Hispanics, indicating Individual 5 matches two Filipino in Saipan or elsewhere in the Pacifi c, likely Native American ancestry, American casualties, as expected given it is possible that they belong to a which in turn, when recovered from the distribution of B4a1a. The group Filipino American or, less likely, North Korea and Papua New Guinea, of casualties matching individual 7, African American service member. It is indicates that the remains most likely despite the variety of racial categories, much more likely, however, that they belong to US service members. are all of Native Hawaiian maternal belong to an indigenous individual. However, we can actually draw ancestry. Given the population history By contrast, remains recovered on the even more specifi c conclusions from a of Hawaii, it is perfectly possible mainland of Asia or in another combat careful examination of the literature. In that the CPD Hispanic who matches theater probably belong to a US service the FRS database, sequence 1 matches this sequence is in fact also of Native member of uncertain ancestry. three Puerto Ricans, and Martínez- Hawaiian maternal descent. Individual Cruzado (2010) has identifi ed it as 6 also matches three casualties of Haplogroup C one of the founding Native American Native Hawaiian maternal ancestry lineages in Puerto Rico. As many and one of uncertain Asian ancestry. Haplogroup C is a branch of US troops of Puerto Rican ancestry However, the last match is to an African haplogroup M distinguished by did serve in Korea, it is likely that American whose maternal lineage can 16223T-16298C-249del (which it these remains represent one of them. be traced to the late nineteenth century shares with haplogroup Z), and 16327T. Sequence 2 is consistent with Mexican in Tennessee and is very unlikely to C1, the primary Native American American examples (Kumar et al. have any Hawaiian ancestry. branch, is further distinguished by 2011), suggesting that that origin is So what accounts for the 16325C-290del-291del. C and its sister more likely. “Polynesian motif” appearing in an Z are largely restricted to northeast Individual 3 presents a very African American? In fact, several Asia, although some branches reach unusual sequence. It bears the C1 African American data sets include further afi eld (Table 2; Derenko et al. motif, but the other fi ve polymorphisms sequences belonging to B4a1a, as well 2010; Ebenesersdóttir et al. 2011). Five are absent from all published examples. as the Southeast Asian haplogroup F3b cases from the sister haplogroups serve Given the number of polymorphisms, Table 6. Haplogroup C Examples Individual 1 2 3 4 5 6 Recovery North Korea Papua New Guinea Saipan North Korea North Korea Hawaii location 16086C, 16183C, 16131C, 16164G, 16189C, 16223T, 16207C, 16223T, 16093C, 16223T, 16185T, 16223T, 16298C, 16325C, 16223T, 16298C, 16278T, 16298C, 16224A, 16234T, 16234T, 16288C, 16260T, 16298C, Sequence 16327T, 215G, 16327T, 16357C, 16325C, 16327T, 16298C, 16325C, 16298C, 16327T, 16301Y, 16362C, 249del, 290-291del 204Y, 207A, 249del 143A, 249del, 290- 16327T, 249del, 249del 151T, 249del 291del 290-291del Haplogroup C1b4 C1c1 C1 C5c1 C4c1? Z Northeast Asian/ Asian/Native Ancestry Native American Native American East European East Asian Native American? American? N/10428 7 Hispanic 2 Hispanic 0 1 Caucasian 0 0 1 White (Hispanic FRS matches 3 Puerto Rican None 1 White 1 American Indian None surname) April 2015 ISSN 2157-118X 9 FORENSIC SCI SEM Alexander F Christensen. Sequence, haplotype, and ancestry. FORENSIC SCI SEM, 2015, 5(1): 1–14. this sequence appears to represent C4a, best documented from South Haplogroup D an old lineage within C1 that has not Asia (Chandrasekhar et al. 2009). been sampled in published studies. It However, the sequence also matches a Haplogroup D is one of the is certainly possible that it is indeed casualty of Native American ancestry. largest clades within haplogroup Native American; it is also possible Given this match, it is possible that M and is found across Asia and that it is Northeast Asian, belonging the sequence actually falls within the the Americas (Achilli et al. 2008; either to C1a with a subsequent loss of recently defi ned C4c, which has been Derenko et al. 2010). Its basic HVR the 16356C that marks that haplogroup documented from multiple Native profi le is 16223T-16362C, which or to a yet undescribed clade. As these Americans and represents an additional unfortunately is also the root profi le remains are from a Pacifi c island, the founding lineage in the hemisphere for multiple other haplogroups, such former option would be consistent with (Hooshiar Kashani et al. 2012). Several as M6, M9 (which includes E), and G. US ancestry, the latter with Japanese of the sequences that fall within this As illustrated previously, some HVR ancestry. haplogroup lack any distinctive HVR profi les within D also match even more Individual 4 is placed within mutations. distantly related phylogenetic twigs. C5 by the 16288C and then C5c1 Individual 6 was recovered on As a result, it is often not possible to by 16093C-16234T. Given C5’s the island of Oahu and is included locate a related sequence precisely distribution in Northeast Asia and the here to show the difference between within the phylogeny. In some cases the recovery location of these remains in haplogroups C and Z. Although sequence may be useful even without North Korea, it would be logical to recovered from a context where a a more precise cladistic position. For assume a Korean or Chinese ancestry. prehistoric Native Hawaiian interment instance, the basic D profi le is found in However, the sequence also happens to was a possibility, the mtDNA sequence approximately 4% of both the Japanese match one White US service member. obtained is inconsistent with that origin, and Korean populations (Jin et al. 2006; Further investigation reveals that C5c1 since haplogroup B4a1a1a would be Sekiguchi et al. 2008). In a forensic is indeed an Eastern European branch, expected in that case. Haplogroup situation where there is a high prior fi rst defi ned from Polish individuals, Z is quite rare and therefore poorly probability of encountering remains and owes its Asian maternal ancestry to documented, so no identical sequences of either of those ancestries, such as a a settler from the steppes millennia ago were encountered in the literature. recovery on a Pacifi c island, this profi le (Derenko et al. 2010). Without the FRS However, very similar examples may be suffi cient to determine the match, it would have been easy to set have been reported from Guangdong, disposition of remains without a more these remains to the side as not from China (Chen et al. 2008; Wang et al. exact phylogeny. the United States and therefore not 2010). The remains likely represent Table 7 provides six casework identifi able by the CIL. a nineteenth- or twentieth-century examples of haplogroup D sequences. Individual 5 provides a similar descendant of East Asian immigrants; The fi rst three were all recovered in case. The sequence lacks the diagnostic depending upon the recovery context, North Korea. Individuals 1 and 2 each markers of C1 and C5, which leaves C4 they could be from either a historical exhibit the 16325C that marks D1. Like and C7 as phylogenetic possibilities, cemetery or a more recent missing A2, D1 has a fairly broad and shallow since neither of these clades is defi ned person case. phylogeny, and it can be diffi cult by HVR mutations. The 16357C to place sequences more precisely. indicates a possible placement within Individual 1’s 16274A-16368C Table 7. Haplogroup D Examples Individual 1 2 3 4 5 6 Recovery North Korea North Korea North Korea Vietnam Papua New Guinea Papua New Guinea location 16093C, 16209C, 16164G, 16172C, 16223T, 16274A, 16223T, 16274A, 16147T, 16223T, 16129A, 16223T, 16182C, 16183C, 16295T, 16325C, 16129A, 16223T, Sequence 16325C, 16362C, 16325C, 150T, 16243C, 16362C, 16189C, 16223T, 16362C, 189G, 16362C, 152C 16368C 152C, 185A, 489C 151T, 152C 16266T, 16362C, 195C, 203A, 204C, 150T 228A, 298T, 325T Haplogroup D1i D1? D4g2a D4a D4a D5a2a1 Ancestry Native American Native American? East Asian East Asian East Asian East Asian N/10428 4 Hispanic 0 0 10 Asian 0 8 Asian 2 White (1 also 1 White (with FRS matches None None None None called Mexican) Spanish surname) 10 Forensic Science Seminar Volume 5 number 1