A New Tool for Probabilistic Assessment of MPS Data Associated with mtDNA Mixtures.

Mitochondrial (mt) DNA plays an important role in the fields of forensic and clinical genetics, molecular anthropology, and population genetics, with mixture interpretation being of particular interest in medical and forensic genetics. The high copy number, haploid state (only a single haplotype contributed per individual), high mutation rate, and well-known phylogeny of mtDNA, makes it an attractive marker for mixture deconvolution in damaged and low quantity samples of all types. Given the desire to deconvolute mtDNA mixtures, the goals of this study were to (1) create a new software, MixtureAceMT™, to deconvolute mtDNA mixtures by assessing and combining two existing software tools, MixtureAce™ and Mixemt, (2) create a dataset of in-silico MPS mixtures from whole mitogenome haplotypes representing a diverse set of population groups, and consisting of two and three contributors at different dilution ratios, and (3) since amplicon targeted sequencing is desirable, and is a commonly used approach in forensic laboratories, create biological mixture data associated with two amplification kits: PowerSeq™ Whole Genome Mito (Promega™, Madison, WI, USA) and Precision ID mtDNA Whole Genome Panel (Thermo Fisher Scientific by AB™, Waltham, MA, USA) to further validate the software for use in forensic laboratories. MixtureAceMT™ provides a user-friendly interface while reducing confounding features such as NUMTs and noise, reducing traditionally prohibitive processing times. The new software was able to detect the correct contributing haplogroups and closely estimate contributor proportions in sequencing data generated from small amplicons for mixtures with minor contributions of ≥5%. A challenge of mixture deconvolution using small amplicon sequencing is the potential generation of spurious haplogroups resulting from private mutations that differ from Phylotree. MixtureAceMT™ was able to resolve these additional haplogroups by including known haplotype/s in the evaluation. In addition, for some samples, the inclusion of known haplotypes was also able to resolve trace contributors (minor contribution 1-2%), which remain challenging to resolve even with deep sequencing.

Recovery of mtDNA from unfired metallic ammunition components with an assessment of sequence profile quality and DNA damage through MPS analysis.

Recovery of suitable amounts of quality DNA from copper and brass surfaces, like those encountered in ammunition, has been a challenge for the forensic community. The ability of copper ions to rapidly facilitate oxidative damage leading to fragmentation of DNA significantly reduces the pool of templates for PCR amplification. We compared two methods for recovering mitochondrial (mt) DNA from the surface of unfired copper projectiles, brass casings, and aluminum casings, and found that using a cotton swab moistened with 0.5M EDTA was the favored approach, especially when the metallic surface was etched. Degradation was significantly higher for DNA samples recovered from copper and brass surfaces, when compared to aluminum. Massively parallel sequencing (MPS) of the control region, using the PowerSeq™ CRM Nested System kit and the Illumina MiSeq instrument, produced full haplotypes for aluminum samples regardless of the method used to deposit or collect DNA, while less than 60% of the copper and brass samples produced partial or full profile information. Touch DNA collected from copper and brass samples produced higher rates of partial or full MPS profile information (∼88-96%), while collection with 0.5M EDTA produced better results than when collection was performed with water; average of ∼70% versus ∼47%. While MPS data was not impacted by noise in the sequencing process, a higher than expected rate of noise was observed, potentially due to an increase in low-level damage lesions. Noise patterns were strikingly different when compared to control data, suggesting that noisy sites may be predictable when testing samples with high levels of oxidative damage. Library preparation was a poor predictor of MPS data quality, as a large percentage of reads did not align with the reference genome. This may impact the number of samples that can be run when a deep-coverage MPS approach is being considered for analysis of mtDNA heteroplasmy. Overall, when applying an MPS approach to the analysis of mtDNA recovered from ammunition, results are expected from touch DNA, will be limited for copper and brass components when the DNA is exposed to an aqueous environment, and DNA degradation will be accelerated when DNA comes in contact with copper or brass surfaces. Practitioners should consider collecting DNA from metallic surfaces with 0.5M EDTA, as this will maximize yield and mitigate degradation. The results of this study directly impact MPS analysis of minor mtDNA sequence variants from metallic surfaces, and are particularly relevant to forensic investigations.

Evaluation of GeneMarker® HTS for improved alignment of mtDNA MPS data, haplotype determination, and heteroplasmy assessment.

Existing software has not allowed for effective alignment of mitochondrial (mt) DNA sequence data generated using a massively parallel sequencing (MPS) approach, combined with the ability to perform a detailed assessment of the data. The regions of sequence that are typically difficult to align are homopolymeric stretches, isolated patterns of SNPs (single nucleotide polymorphisms), and INDELs (insertions/deletions). A custom software solution, GeneMarker® HTS, was developed and evaluated to address these limitations, and to provide a user-friendly interface for forensic practitioners and others interested in mtDNA analysis of MPS data. GeneMarker® HTS generates an exportable consensus mtDNA sequence that produces phylogenetically correct SNP and INDEL calls using a customizable motif-based alignment algorithm. Sequence data from 500 individuals, with various alignment asymmetries and levels of heteroplasmy, were used to assess the software. Accuracy in producing mtDNA haplotypes, the ability to correctly identify low-level heteroplasmic sequence variants, and the user-based features of the software were evaluated. Analyzed sequences yielded correct mtDNA haplotypes, and heteroplasmic variants were properly identified with minimal manual interpretation. The software offers numerous user-defined parameters for filtering the data that address the interests of researchers and practitioners, and provides multiple options for viewing and navigating through the data.

Development and assessment of an optimized next-generation DNA sequencing approach for the mtgenome using the Illumina MiSeq.

The development of molecular tools to detect and report mitochondrial DNA (mtDNA) heteroplasmy will increase the discrimination potential of the testing method when applied to forensic cases. The inherent limitations of the current state-of-the-art, Sanger-based sequencing, including constrictions in speed, throughput, and resolution, have hindered progress in this area. With the advent of next-generation sequencing (NGS) approaches, it is now possible to clearly identify heteroplasmic variants, and at a much lower level than previously possible. However, in order to bring these approaches into forensic laboratories and subsequently as accepted scientific information in a court of law, validated methods will be required to produce and analyze NGS data. We report here on the development of an optimized approach to NGS analysis for the mtDNA genome (mtgenome) using the Illumina MiSeq instrument. This optimized protocol allows for the production of more than 5 gigabases of mtDNA sequence per run, sufficient for detection and reliable reporting of minor heteroplasmic variants down to approximately 0.5-1.0% when multiplexing twelve samples. Depending on sample throughput needs, sequence coverage rates can be set at various levels, but were optimized here for at least 5000 reads. In addition, analysis parameters are provided for a commercially available software package that identify the highest quality sequencing reads and effectively filter out sequencing-based noise. With this method it will be possible to measure the rates of low-level heteroplasmy across the mtgenome, evaluate the transmission of heteroplasmy between the generations of maternal lineages, and assess the drift of variant sequences between different tissue types within an individual.