As submitted by the proposer: Aging human remains using skeletal and dental features is common practice in forensic science, and is often a first step to building a biological profile of a victim. However, fewer methods are available for building the profile of a suspect; even if a DNA sample can be isolated from a crime scene, the utility of the genetic data for identification requires that the individual or a near relative is already present in a forensic database. We propose to develop a method of age estimation of both victims and suspects utilizing DNA methylation, an epigenetic modification that involves the addition of a methyl group to a cytosine (C) in a CG dinucleotide context. It has been demonstrated that patterns of methylation exhibit significant, genome-wide changes throughout individual life, although the effects of ethnicity and tissue type on these patterns is not yet fully understood or appreciated.
We will use currently available methylation data from saliva and blood from Europeans, Hispanic-Latinos, and Africans, and will generate new data from saliva from Africans (n = 35) and bone from Americans (n = 35). By using methylation data both from genetically diverse human groups and multiple tissue types, I will design epigenetic age prediction algorithms that can be used on multiple types of biological traces that are left at a crime scene. By using commercially available array technology, I will assay the methylation state of over 450,000 distinct sites in the genome.
This data will then be used to identify sites of methylation that exhibit statistically significant age-related changes. A collection of sites that vary reliably with age will be used to develop the prediction models. The statistical model will be trained on half of the samples and tested on the remainder of the samples.
The resulting data will include genome-wide methylation levels for all individuals and associated variables, such as age and ancestry. This research will result in optimized sets of sites and statistical models that will accurately predict age using DNA methylation data as input. This project will build on previous research and epigenetic age models to address the problem of forensic age prediction in cosmopolitan jurisdictions in the United States by 1) considering a wider range of human genetic diversity and 2) generating a new dataset of bone methylation which will be used to accurately age individuals from skeletal remains.
This project contains a research and/or development component, as defined in the applicable law, and complies with Part 200 Uniform Requirements 2 CFR 200.210(a) (14). nca/ncf
