U.S. flag

An official website of the United States government, Department of Justice.

Forensic Body Fluid Identification Using Microbiome Signature Attribution

NCJ Number
256087
Date Published
September 2020
Author(s)
Baneshwar Singh Ph.D.; Andrea Publow, MFA, CRA
Agencies
NIJ-Sponsored
Grant Number(s)
2016-DN-BX-0151
Annotation

This is the Final Draft Summary Report of a research project that developed a novel non-human DNA (microbiome) – based method for the identification of the major forensically relevant human biological fluids.

Abstract

ABST Project results indicate that all major body fluid samples can be distinguished and identified with an overall accuracy of 89.9 percent using this technique. Vaginal and menstrual secretions were indistinguishable from each other, and they were classified together as “female intimate samples.”  Developmental validation of this method indicated that a reliable microbial profile could be obtained from a sample that has bacterial DNA quantity of at least 5pg. The newly developed method is also robust to many common environmental contaminants. Overall, the proposed method is fast (no additional steps are needed, and one test can identify all major body fluids), accurate, and easily added into a forensic high throughput sequencing (HTS) panel. In testing the proposed method, 1,160 biological samples (semen, vaginal secretions, menstrual secretions, saliva, feces, urine, and venous blood) were collected and preserved using methods common in forensic laboratories for evidence collection. Except for urine, DNA from all other samples were extracted using the QIAamg DNA investigator Kit standard forensic casework sample protocol on the QIAcube robotic workstation. DNA from urine was extracted using the QIAamp DNA Micro Kit. Variable region four (V4) of 16S ribosomal DNA (16S rDNA) was amplified using a dual-index strategy and then sequenced on the MiSeq FGx sequencing platform using the MiSeq Reagent Kit v2 (500 cycles) and following the manufacturer’s protocol. Sequence data were analyzed using mothur version 1.39.1 and R version 3.4.0. A novel “support vector machine” technique was implemented in R to estimate the accuracy. 13 figures, 4 tables, and 31 references         

Date Created: February 11, 2021