NIST Interlaboratory Mixture Interpretation Study 2013 (MIX13)
[Download Data] [Presentation on Study Goals] [Responses to Frequently Asked Questions]
Introduction
The MIX13 interlaboratory mixture study involves only data interpretation. We are providing electronic data generated at NIST and Boston University (BU) to eliminate the variability in differences due to instrument sensitivities and PCR amplification. Mixtures representing five different case scenarios have been generated and are available as .fsa files that can be downloaded below.
Purpose
The last NIST interlaboratory study regarding mixture interpretation was conducted in 2005 (MIX05). In the ensuing years, a great deal of effort has focused on improving DNA mixture interpretation around the world. In the U.S., the SWGDAM group published a set of guidelines in 2010 for autosomal STR interpretation. In this study we aim (1) to evaluate the current “lay of the land” regarding STR mixture interpretation across the community (2) to measure consistency in mixture interpretation across the U.S. after the publication of the 2010 SWGDAM guidelines and (3) to learn where future training and research could help improve mixture interpretation and reporting.
Study Details
Five cases are provided each with an evidentiary sample file (a mixture of at least one suspect and one victim), and depending on the scenario, a victim reference profile, suspect(s) profiles and other known references. We have generated .fsa files on an ABI Genetic Analyzer 3130xl using either PowerPlex16HS or Identifiler (BU)/IdentifilerPlus (NIST) kits. Allelic ladders, positive and negative controls are also provided.
Case #1 – represents sexual assault evidence from the sperm fraction of a vaginal swab. The genotype of the female victim and a male suspect are provided for comparison. Data available in Identifiler Plus and PP16HS.
Case #2 – represents evidence swabbed from the handle of a handgun retrieved outside the front door of a convenience store robbery/homicide. Reference samples from four gang members are provided for comparison. Data generated by BU in either Identifiler or PP16HS. Note: BU used the GS600 LIZ size standard for this Identifiler sample – all other examples used the GS500 LIZ size standard.
Case #3 – represents sexual assault evidence from the sperm fraction of a vaginal swab. The genotype of the female victim, a male consensual partner and two suspects (one of whom is the brother of the consensual partner) are provided for comparison. Data available in Identifiler Plus and PP16HS.
Case #4 – represents bite-mark evidence from a victim assaulted at a bus stop. The genotype of the female victim and a male suspect are provided for comparison. Data available in Identifiler Plus and PP16HS.
Case #5 – represents a ski-mask recovered at a bank robbery where a gang of at least four individuals have committed several robberies over the last several months. Three suspect genotypes are provided for comparison. Data available in Identifiler Plus and PP16HS.
Data Interpretation
An important purpose of this study is to develop an understanding on how laboratories are interpreting mixtures using their SOPs developed from their validation studies and the 2010 SWGDAM guidelines. Because of the variation inherent in a study where the data was generated on instrumentation in one lab but analyzed by protocols developed by validation studies in another lab, each lab may wish to take two different approaches to the interpretation of the data.
(1) Each lab is welcome to use the NIST and BU established thresholds for interpretation (NIST: AT = 50 RFU, ST = 150 RFU; BU: AT = 30 RFU, ST = 150 RFU)
(2) Each lab is welcome to use their own thresholds and analysis parameters according to their own guidelines.
Since we expect to find variation from individual laboratories having differing AT and ST values, we ask for Case #4 (bite-mark example) that all labs analyze this example using the NIST threshold values (again, this may be in addition to your own analysis with your guidelines).
Finally, the success of this study depends on two major factors – lab participation and data generated from the analysis. We tried to develop realistic scenarios in our examples, and realize that your options are limited for generating “better” profiles via re-extraction, re-amplification, and re-injection. Please make every effort to analyze these cases “as is” instead of abstaining because, “I would re-inject this example.”
Prior to October 4, 2013 (if you are a U.S. public laboratory), we would like to receive the following information:
1) For U.S. public laboratories, please return only one official TL lab response for each of the five cases. For all others, your participation and the data from this interlaboratory study will be a crucial component for this evaluation. We would like to encourage anyone to participate in this study as an opportunity to learn and improve. This study is open to the international community. We plan to conclude collecting data by December 31, 2013.
2) Report the results as though they were from a real case and include a brief explanation as to why conclusions were reached in each case scenario. Please summarize the alleles present in the mixture and the perpetrator(s) alleles in each “case” as they might be presented in court or in a laboratory report. Please indicate the AT and ST parameters used for the analysis.
3) If a statistical analysis is performed, please present this and indicate how this was determined. Please include the source of the allele frequencies used to make the calculation. Also indicate how your statistical calculation was performed (e.g. Popstats, in-house Excel, etc…)
4) If you are interpreting the mixtures with your own laboratory mixture interpretation guidelines it would be useful to include a copy if possible.
As with previous NIST interlaboratory studies, each participant will receive (after providing results) a certificate of study participation and a summary report of the findings. Participants will be provided a review copy of the final report prior to its submission to a peer-reviewed journal. Results of individual laboratory or analyst performance will be provided directly to the lab upon request and overall study metrics will be reported anonymously for presentations and publications.
Thank you for your participation. Please contact me if you have any questions.
Michael Coble
Forensic Biologist
NIST - Applied Genetics Group
100 Bureau Drive MS 8314
Gaithersburg, MD 20899-8314
Phone: 301-975-4330
Email: michael.coble@nist.gov
Download MIX13 Data: (click on desired .fsa file, then save to your computer in desired directory)
Reference Profiles for Comparison Purposes
(Excel file with different suspects for each case in individual tabs)
|
Identifiler Plus Data |
PowerPlex 16 HS Data |
|||||||||
| Zip file containing all samples from the 5 cases | Zip file containing all samples from the 5 cases | |||||||||
| Case 1 | Case 1 | |||||||||
|
|
|||||||||
| Case 2 | Case 2 | |||||||||
|
|
|||||||||
| Case 3 | Case 3 | |||||||||
|
|
|||||||||
| Case 4 | Case 4 | |||||||||
|
|
|||||||||
| Case 5 | Case 5 | |||||||||
|
|
|||||||||
|
|
*PP16HS Users: In our attempt to make the mixtures as consistent as possible with the labs performing Identifiler amplification, we found that our positive control was diluted below our stochastic threshold. We have not included this in the data folder. If you have any questions, please let me know. |
Last Updated: 08/26/2013