Darrell O. Ricke

Robust detection of individual forensic profiles in DNA mixtures

For a forensic identification method to be admissible in international courts, the probability of false match must be quantified. For comparison of individuals against complex mixtures using a panel of single nucleotide polymorphisms (SNPs), the probability of a random man not excluded, P(RMNE) is one admissible standard. While the P(RMNE) of SNP alleles has been previously studied, it remains to be rigorously defined and calculated for experimentally genotyped mixtures. In this report, exact P(RMNE) values were calculated for a range of complex mixtures, verified with Monte Carlo simulations, and compared alongside experimentally determined detection probabilities.

In vivo Monitoring of Transcriptional Dynamics After Lower-Limb Muscle Injury Enables Quantitative Classification of Healing

Traumatic lower-limb musculoskeletal injuries are pervasive amongst athletes and the military and typically an individual returns to activity prior to fully healing, increasing a predisposition for additional injuries and chronic pain. Monitoring healing progression after a musculoskeletal injury typically involves different types of imaging but these approaches suffer from several disadvantages. Isolating and profiling transcripts from the injured site would abrogate these shortcomings and provide enumerative insights into the regenerative potential of an individual’s muscle after injury. In this study, a traumatic injury was administered to a mouse model and healing progression was examined from 3 hours to 1 month using high-throughput RNA-Sequencing (RNA-Seq). Comprehensive dissection of the genome-wide datasets revealed the injured site to be a dynamic, heterogeneous environment composed of multiple cell types and thousands of genes undergoing significant expression changes in highly regulated networks. Four independent approaches were used to determine the set of genes, isoforms, and genetic pathways most characteristic of different time points post-injury and two novel approaches were developed to classify injured tissues at different time points. These results highlight the possibility to quantitatively track healing progression in situ via transcript profiling using high- throughput sequencing.

Genetic sequence matching using D4M big data approaches

Recent technological advances in Next Generation Sequencing tools have led to increasing speeds of DNA sample collection, preparation, and sequencing. One instrument can produce over 600 Gb of genetic sequence data in a single run. This creates new opportunities to efficiently handle the increasing workload. We propose a new method of fast genetic sequence analysis using the Dynamic Distributed Dimensional Data Model (D4M) - an associative array environment for MATLAB developed at MIT Lincoln Laboratory. Based on mathematical and statistical properties, the method leverages big data techniques and the implementation of an Apache Acculumo database to accelerate computations one-hundred fold over other methods. Comparisons of the D4M method with the current gold-standard for sequence analysis, BLAST, show the two are comparable in the alignments they find. This paper will present an overview of the D4M genetic sequence algorithm and statistical comparisons with BLAST.

Sherlock's Toolkit: A forensic DNA analysis system

DNA sequence analysis has multiple forensic applications. The justice system currently uses sizes of STRs as well as mitochondrial DNA (mtDNA) for DNA evidence. With recent advancements in DNA sequencing technologies, inclusion of additional polymorphic loci, including SNPs, enable new useful analyses while maintaining backwards compatibility with STR sizing. Sherlocks Toolkit, developed by MIT Lincoln Laboratory, is an open source, scalable system for the integration and automation of STR and SNP-based analysis for high-throughput sequence data. The toolkit includes modules for a range of kinship, biogeographic ancestry, replicate, and mixture analyses.

Association Between Single Gene Polymorphisms and Bone Biomarkers and Response to Calcium and Vitamin D Supplementation in Young Adults Undergoing Military Training

Initial military training (IMT) is associated with increased stress fracture risk. In prior studies, supplemental calcium (Ca) and vitamin D provided daily throughout IMT reduced stress fracture incidence, suppressed parathyroid hormone (PTH), and improved measures of bone health compared with placebo. Data were analyzed from a randomized, double‐blind, placebo‐controlled trial to determine whether single‐nucleotide polymorphisms (SNPs) in Ca and vitamin D–related genes were associated with circulating biomarkers of bone metabolism in young adults entering IMT, and whether responses to Ca and vitamin D supplementation were modulated by genotype. Associations between SNPs, including vitamin D receptor (VDR), vitamin D binding protein (DBP), and 1‐alpha‐hydroxylase (CYP27B1), and circulating biomarkers were measured in fasting blood samples from volunteers (n = 748) starting IMT. Volunteers were block randomized by race and sex to receive Ca (2000 mg) and vitamin D (1000 IU) or placebo daily throughout Army or Air Force IMT (7 to 9 weeks). Total Ca and vitamin D intakes were calculated as the sum of supplemental intake based on intervention compliance and dietary intake. Relationships between SNPs, Ca, and vitamin D intake tertile and change in biomarkers were evaluated in trial completers (n = 391). At baseline, the minor allele of a DBP SNP (rs7041) was positively associated with both 25OHD (B = 4.46, p = 1.97E‐10) and 1,25(OH)2D3 (B = 9.63, p < 0.001). Combined genetic risk score (GRS) for this SNP and a second SNP in the VDR gene (rs1544410) was inversely associated with baseline 25OHD (r = –0.28, p < 0.001) and response to Ca and vitamin D intake differed by GRS (p < 0.05). In addition, presence of the minor allele of a second VDR SNP (rs2228570) was associated with lower P1NP (B = –4.83, p = 0.04) and osteocalcin (B = –0.59, p = 0.03). These data suggest that VDR and DBP SNPs are associated with 25OHD status and bone turnover and those with the highest GRS require the greatest vitamin D intake to improve 25OHD during IMT.

KinLinks: Software Toolkit for kinship analysis and pedigree generation from HTS datasets

The ability to predict familial relationships from source DNA in multiple samples has a number of forensic and medical applications. Kinship testing of suspect DNA profiles against relatives in a law enforcement database can provide valuable investigative leads, determination of familial relationships can inform immigration decisions, and remains identification can provide closure to families of missing individuals. The proliferation of High-Throughput Sequencing technologies allows for enhanced capabilities to accurately predict familial relationships to the third degree and beyond. KinLinks, developed by MIT Lincoln Laboratory, is a software tool that predicts pairwise relationships and reconstructs kinship pedigrees for multiple input samples using single-nucleotide polymorphism (SNP) profiles. The software has been trained and evaluated on a set of 175 subjects (30,450 pairwise relationships), consisting of three multi-generational families and 52 geographically diverse subjects. Though a panel of 5396 SNPs was selected for kinship prediction, KinLinks is highly modular, allowing for the substitution of expanded SNP panels and additional training models as sequencing capabilities continue to progress. KinLinks builds on the SNP-calling capabilities of Sherlocks Toolkit, and is fully integrated with the Sherlocks Toolkit pipeline.

Divergence Model of Protein Evolution

Analysis of the evolutionary conservation of amino acids is useful for the analysis of sequence variants detected in individuals with regard to possible impact on protein function. Rapid advances in DNA sequencing technologies are enabling affordable access to SNPs, exome sequencing, and whole genome shotgun sequencing. An understanding of the biological processes that have shaped diverging protein sequences aids the interpretation of sequence variants. The divergence model of protein evolution is presented as a general framework for interpreting information available from comparative analysis of protein sequences.

FastID: Extremely fast forensic DNA comparisons

Rapid analysis of DNA forensic samples can have a critical impact on time sensitive investigations. Analysis of forensic DNA samples by massively parallel sequencing is creating the next gold standard for DNA forensic analysis. This technology enables the expansion of forensic profiles from the current 20 short tandem repeat (STR) loci to tens of thousands of single nucleotide polymorphism (SNP) loci. A forensic search scales by the product of the number of loci and the number of profile comparisons. This paper introduces a method (FastID) to address the need for rapid scalable analysis of DNA forensic samples (patent pending)[1]. FastID can search a profile of 2,500 SNP loci against 20 million profiles in 5.08 seconds using a single computational thread on a laptop (Intel i7 4.0 GHz).

Towards an open data framework for body sensor networks supporting bluetooth low energy

Major companies, healthcare professionals, the military, and other scientists and innovators are now sensing that fitness and health data from wearable biosensors will likely provide new discoveries and insights into physiological, cognitive, and emotional health status of an individual. Having the ability to collect, process, and correlate data simultaneously from a set of heterogonous biosensor sources may be a key factor in informing the development of new technologies for reducing health risks, improving health status, and possibly preventing and predicting disease. The challenge in achieving this is getting easy access to heterogeneous data from a set of disparate sensors in a single, integrated wearable monitoring system. Often times, the data recorded by commercial biosensing devices are contained within each manufacturers proprietary platform. Summary data is available for some devices as free downloads or included only in annual premium memberships. Access to raw measurements is generally unavailable, especially from a custom developed application that may include prototype biosensors. In this paper, we explore key ideas on how to leverage the design features of Bluetooth Low Energy to ease the integration of disparate biosensors at the sensor communication layer. This component is intended to fit into a larger, multi-layered, open data framework that can provide additional data management and analytics capabilities for consumers and scientists alike at all the layers of a data access model which is typically employed in a body sensor network system.

Human CODIS STR loci profiling from HTS data

Human DNA identification is currently performed by amplifying a small, defined set of short tandem repeat (STR) loci (e.g. CODIS) and analyzing the size of the alleles present at those loci by capillary electrophoresis. High-throughput DNA sequencing (HTS) could enable the simultaneous analysis of many additional STR and single nucleotide polymorphism (SNP) loci, improving accuracy and discrimination. However, it is necessary to demonstrate that HTS can generate accurate data on the CODIS loci to enable backwards compatibility with the FBI NDIS database. Sequencing can also detect novel polymorphisms within alleles that migrate with identical sizes by capillary electrophoresis, improving allele discrimination, and enhancing human identification analysis. All CODIS alleles from an individual can be amplified in a single, multiplex PCR reaction, and combined with additional barcoded samples prior to sequencing. A computational tool for allele identification from multiplexed sequence data has been developed. With longer-read-length platforms, 99.6% allele calling accuracy can be achieved. In the course of STR sequencing protocol development, 12 novel allele sequences have been identified for multiple loci. Sequencing STR loci combined with SNPs will enable new forensic applications.