Research article| Volume 2, ISSUE 1, P245-247, December 2009

DNA profiling of skeletal samples from the disappeared in Latin America

Published:November 09, 2009DOI:


      The identification of missing persons can be quite challenging; particularly in instances where a significant number of individuals went missing during conflicts from decades past. In these situations, DNA profiling and kinship analysis are necessary tools to aid with the identification of compromised skeletal remains such as those analyzed for the Latin American Initiative for the Identification of the “Disappeared.” In 2008, Bode was awarded contracts for this initiative to perform DNA analysis on hundreds of skeletal remains and family references. Bode's success in obtaining profiles utilizing advanced extraction and purification techniques along with mini-STR technology has, to date, resulted in the identification of greater than 80 individuals.


      1. Introduction

      Over the course of several decades, ranging from the 1960s to the 1990s, approximately 300,000 Latin Americans went missing as a result of internal conflicts throughout Central and South America. In an effort to bring closure to the families of those “disappeared”, organizations such as the Guatemalan Forensic Anthropology Foundation (FAFG), the Peruvian Forensic Anthropology Team (EPAF), and the Argentine Forensic Anthropology Team (EAAF), along with assistance from The Bode Technology Group, Inc. (Bode), and Non-Governmental Organizations such as the Washington Office for Latin America (WOLA), helped institute the Latin American Initiative for the Identification of the “Disappeared” (LIID). Funding for the initiative was provided by the Bureau of Democracy, Human Rights and Labor (DRL) of the US State Department, enabling DNA testing to be performed on the skeletal remains from these missing persons.
      The skeletal remains received from Guatemala, Peru, and Argentina had undergone various levels of environmental insult due to exposure to heat, water, bacteria and soil elements. All of these factors, combined with underground burials for over twenty years and in some cases, co-mingling of samples in mass graves, detrimentally affected the ability to recover intact and uncompromised DNA. A significant proportion of partial genotypes were obtained from the skeletal remains, which posed complexities for kinship analysis with large numbers of familial reference genotypes. This paper outlines challenges that were encountered as well as lessons learned with DNA typing and kinship analysis of human remains related to the LIID project.

      2. Materials and methods

      Skeletal remains were decalcified with EDTA and extracted using the QIAamp® DNA Blood Maxi Kit (QIAGEN, Inc., Valencia, CA) with modifications. DNA extracts were concentrated using Vivacon 2 Ultrafiltration Devices (Sartorius Stedim Biotech S.A, Aubagne, France). DNA yield was measured with either BodeQuant LCN, a real-time human DNA quantitation method developed at Bode Technology, or Quantifiler™ (Applied Biosystems, Foster City, CA). Inhibited or degraded DNA samples were purified pre-PCR via Centri Sep 8 columns (Princeton Separations, Inc., Freehold, NJ). All samples were amplified with the AmpFℓSTR® Identifiler® PCR Amplification Kit for 28 cycles, and select samples were amplified using the AmpFℓSTR® MiniFiler™ PCR Amplification Kit for 30 cycles. Amplification products were separated and detected using the ABI PRISM® 3100 Genetic Analyzer, and analyzed with GeneMapper® ID Software v3.2 (Applied Biosystems).
      Familial reference samples were extracted using either the QIAamp DNA Micro Kit, the BioSprint 96 DNA Blood Kit on the BioSprint 96 robot, or the EZ1 DNA Investigator Kit on the BioRobot EZ1 (QIAGEN). Quantitation, amplification, separation and detection, and analysis were performed using the same methods and parameters as the skeletal samples.
      Kinship analysis was performed utilizing a software program developed by Bode, Bode-Link, which screened each skeletal genotype against all reference genotypes for evidence of relatedness through calculation of paternity (PI) and sibling (SI) kinship indices. Each skeletal sample that was consistent with a reference was further screened against all family units that included the reference, to reduce the number of fortuitous matches. Skeletal genotypes were also screened against each other for intra-skeletal direct matches. Then using Southwest Hispanic population frequencies, the DNA·VIEW™ software was employed to confirm identifications by assigning a probability calculation to each match [

      C.H. Brenner, DNA·VIEW™ 2009 User's Manual 27 June (2009)

      ]. Only matches with a posterior probability of greater than 99.9% were officially reported as identifications [
      • Ritter N.
      Identifying remains: lessons learned from 9/11.

      3. Results and discussion

      Similar challenges were posed by the circumstances surrounding DNA profiling for each LIID project. Many bone samples were inhibited to the point where no DNA was detected upon quantitation. After traditional dilution procedures failed to yield STR data, Bode employed the use of additional purification via Centri Sep prior to amplification. The combination of additional purification and the use of mini-STRs was instrumental in obtaining genotypes from samples where analysis with conventional STRs did not yield a full profile. Fig. 1 illustrates comparative data for each of the LIID projects for only Identifiler analysis versus combined Identifiler and MiniFiler analysis, where obtained profiles had greater than 10 complete loci.
      Figure thumbnail gr1
      Fig. 1A comparison of skeletal DNA profiles (with 10 or more loci) obtained using only Identifiler analysis versus composite profiles obtained from the combination of Identifiler and MiniFiler analyses.
      Differing challenges were observed with kinship analysis for each project, which are described along with project specific data below.

      3.1 Guatemalan forensic anthropology foundation (FAFG)

      DNA analysis on skeletal and familial reference samples from the FAFG was completed as of 31 December 2008. Bode analyzed 143 bones and teeth and 451 reference samples. DNA profiling of the skeletal samples for this project was exceptionally challenging as they appeared to be saturated with soil particles. Even after extensive cleaning apparent soil contaminants still hindered flow through the column membrane during extraction, requiring additional purification for a greater proportion of samples.
      Kinship analysis was not performed at Bode for this project; however, the FAFG has confirmed that they were immediately able to make 12 identifications upon receipt of data delivered by Bode with further identifications being reviewed.

      3.2 Peruvian forensic anthropology team (EPAF)

      The skeletal samples for this project were associated with three specific cases, where each case represented one or more mass grave(s) from a certain village. Skeletal samples and references were only compared within each isolated case. Bode analyzed 123 bones and teeth and 150 references samples for the EPAF.
      Thorough genealogy information was provided by the EPAF to aid in kinship analysis; however, the matching process was still extremely complex because many of the missing were to serve as references for other missing individuals. Using Bode-Link, presumptive identifications were generated from the strongest likelihood ratios. Bode was able to re-associate families by comparing familial references to skeletal samples, and by comparing any newly identified skeletal profiles to those still unidentified. All kinship matches meeting the posterior probability reporting threshold were checked by the EPAF against ante-mortem and anthropological information before being officially reported. Thus far, 25 official match reports were issued by Bode for the EPAF project leading to the identification of over 30 individuals.

      3.3 Argentine forensic anthropology team (EAAF)

      Bode processed approximately 800 skeletal remains from 600 individuals and over 5200 family reference samples during the initial phase of testing for the EAAF. In addition to the profiles obtained by Bode, the EAAF also submitted about 300 previously obtained familial reference profiles to be compared to all of the skeletal genotypes.
      This project required comparison of an exceptionally large number of skeletal and reference DNA profiles. Bode-Link was instrumental in searching the thousands of profiles obtained for possible kinship matches. Kinship information provided by the EAAF was used to determine probability of these matches. A potential identifications with a PI or SI likelihood ratio greater than 500 were submitted to EAAF for review. Only those meeting matching criteria were checked for relatedness with DNA·VIEW™ in both the Southwest Hispanic population and the Argentinean population [
      • Borosky A.
      • Catelli L.
      • Vullo C.
      Analysis of 17 STR loci in different provinces of Argentina.
      ]. Matches above the posterior probability threshold were reported to the EAAF for confirmation before official release. Overall, the work Bode performed for the EAAF has resulted in the identification of 42 individuals thus far, with potential additional identifications in the process of being confirmed.

      4. Conclusions

      Advanced extraction techniques, pre-PCR purification and amplification using a combination of traditional and mini-STR analysis methods were found to be essential tools in obtaining sufficient DNA data for kinship comparison. These tools allowed Bode to overcome the challenges associated with processing compromised skeletal remains and ultimately obtain a greater number of loci for each DNA profile. Furthermore, software for screening large numbers of unknown profiles against familial reference profiles is essential in detecting potential matches when attempting to perform kinship analysis on a large scale. To address this need Bode developed the Bode-Link software, which, when combined with DNA·VIEW™, enabled the identification of greater than 80 individuals. The difficulties encountered while processing and analyzing skeletal samples associated with the LIID initiative have taught everybody involved many valuable lessons, which will undoubtedly increase the efficiency and success of further testing. Additional funding is being sought to continue the realized success of this mission, as many individuals have yet to be identified.

      Role of funding

      Funding for the LIID initiative was provided by the Bureau of Democracy, Human Rights and Labor (DRL) of the US State Department. However, the DRL had no involvement in the collection or processing of samples for this initiative.

      Conflict of interest



      We would like to acknowledge Emily Snyder and Caroline Chen for their instrumental role in obtaining DNA profiles from the skeletal remains for this project, Erin Sweeney for leading the efforts to develop DNA profiles from the familial reference samples, and Mike Cariola and Edwin Huffine for guidance and editing assistance in development of this article.


      1. C.H. Brenner, DNA·VIEW™ 2009 User's Manual 27 June (2009)

        • Ritter N.
        Identifying remains: lessons learned from 9/11.
        NIJ J. 2007; 256
        • Borosky A.
        • Catelli L.
        • Vullo C.
        Analysis of 17 STR loci in different provinces of Argentina.
        Forensic Sci. Int. Gene. 2009; 3: e93-e95