NIJ Awards Over $11 Million to Support Forensic Science Research and Development in 2022

Evaluating scenarios to explain why a suspect’s DNA is found at a crime scene can help speak to activity levels that provides value to the criminal justice system, especially when backed by empirical data. The goal of the proposed research is to create a ground truth sample, the domesticated fingerprint and hand that will eliminate variation in DNA quantity when used to make touch DNA samples. Specific DNA transfer events and chains of events will be evaluated by measuring DNA transfer and recovery. The resulting data will be used to establish a “bullpen of transfer events,” each supported by empirical transfer and recovery data. This would, in turn, allow forensic scientists to objectively evaluate the potential for DNA profiles to be generated by a variety of potential scenarios. 

Familial searches using 13-20+ short tandem repeats (STRs) in a DNA profile are sufficient to provide linkages to full siblings and parent-offspring pairs. However, more distant relationships cannot be detected without additional genetic markers. This has led to the prominence of genetic genealogical and related methods being applied to unsolved forensic investigations. Many factors can impact the performance of forensic genetic genealogy, including, but not limited to, input data quality, database size, and the number and genetic distance of cousin matches within the database. This study aims to assess the effectiveness of forensic genetic genealogy techniques for persons identified as Black, Hispanic, and Native American. Performance for de-identified samples with self-reported ancestry will be evaluated and human skeletal samples will be evaluated from individuals estimated to be Black, Indigenous, or Persons of Color (BIPOC) based on anthropological assessment. Finally, the efficacy of genetic genealogy methods for use in ways that would be inclusive of BIPOC persons will be tested.

Familial searches using 13-20+ short tandem repeats (STRs) in a DNA profile are sufficient to provide linkages to full siblings and parent-offspring pairs. However, more distant relationships cannot be detected without additional genetic markers. This has led to the prominence of genetic genealogical and related methods being applied to unsolved forensic investigations. Many factors can impact the performance of forensic genetic genealogy, including, but not limited to, input data quality, database size, and the number and genetic distance of cousin matches within the database. This study aims to assess the effectiveness of forensic genetic genealogy techniques for persons identified as Black, Hispanic, and Native American. Performance for de-identified samples with self-reported ancestry will be evaluated and human skeletal samples will be evaluated from individuals estimated to be Black, Indigenous, or Persons of Color (BIPOC) based on anthropological assessment. Finally, the efficacy of genetic genealogy methods for use in ways that would be inclusive of BIPOC persons will be tested.

The legalization of hemp-type cannabis and corresponding definition of marijuana as containing >0.3% THC has dramatically increased the workload in forensic laboratories, which must analyze and quantify the THC content of incoming cannabis samples so that seized material can be appropriately designated. This project seeks to develop a novel screening method to differentiate between hemp and marijuana by assessing the binding interactions of cannabinoids with metals using computational calculations for prediction of those that will yield differential color, UV-Vis, or fluorescence characteristics. These high-level quantum calculations will inform the experimental assessment of the differential color and fluorescence characteristics of cannabinoid/metal binding (including THC/Metal and CBD/Metal pairs).

Marijuana is the most abused drug across the world and is known to severely impair skills needed for safe driving. Identifying drivers under the influence of marijuana is important for ensuring public safety, and there is an urgent need for a device, similar to the alcohol breath analyzer, that can be used by law enforcement officers to accurately and non-invasively quantify THC levels in roadside scenarios. This project focuses on developing an aptamer-modified paper electrochemical device that can be used by law enforcement officers to easily quantify THC levels in the oral fluid of drivers. The planned device should require only ~10-20 microliters of sample and provide results in a digital readout within seconds.

Footwear evidence, like shoe prints and impressions left at crime scenes, is used less frequently in the United States than in other countries. One factor limiting the use of this evidence in criminal investigations is the lack of reference data on the footwear make and model and outsole patterns currently circulating in the U.S. This project proposes to develop a mobile app for capturing footwear images and tagging them with metadata. The researchers will then initiate a crowdsourcing campaign to engage the public in collecting a footwear reference dataset. Finally, they plan to evaluate and make recommendations on the applicability of this citizen science model to other forensic science disciplines.

The interpretation of fire damage for investigation of possible arson is largely a qualitative, subjective process performed by human investigators. Experimental research under controlled conditions can help investigators understand the relationship between fire conditions and the resulting damage patterns, but most prior research has been done on unfinished drywall without regard to painting or finishing. This project will systematically examine the influence of a variety of common finishes and paints on the patterns resulting from fire damage. The researchers will then apply computer vision and machine learning methods to the data to build objective classification tools for use by fire investigators.

Estimating biological profiles (i.e., sex, population affinity, age, and stature) from skeletal remains is an essential task of forensic anthropology and helps in matching skeletal remains of an unknown individual to a reported missing person. The overall objective of this project is to develop deep learning models to estimate the skeletal characteristics of like groups of people (the population affinity) using craniometric data. The researchers will implement their models into a web service for use in population affinity estimation.

Through prior NIJ funding over the last decade, the research team at Colorado State University has demonstrated the power of microbiome-based estimates of postmortem interval with a mouse model system, and then translated those estimates to human cadavers utilizing multiple U.S. anthropological research facilities. The previous research resulted in a postmortem microbiome database. Under the current project, the researchers aim to: 1) expand the postmortem microbiome database by collecting samples from facilities in a climate type not yet represented in the database, and 2) validate the postmortem microbiome database with an independent test set.