Biology subjects

Glotov, A. S.

Publications and source records attributed to Glotov, A. S..

3 recordsLinked to original sources

bioRxiv2021

Systematic benchmark of state-of-the-art variant calling pipelines identifies major factors affecting accuracy of coding sequence variant discovery

Accurate variant detection in the coding regions of the human genome is a key requirement for molecular diagnostics of Mendelian disorders. Efficiency of variant discovery from nextgeneration sequencing (NGS) data depends on multiple factors, including reproducible coverage biases of NGS methods and the performance of read alignment and variant calling software. In this work, we systematically evaluated the performance of 4 popular short read aligners (Bowtie2, BWA, Isaac, and Novoalign) and 6 variant calling and filtering methods (based on DeepVariant, Genome Analysis ToolKit (GATK), FreeBayes, and Strelka2) using a set of 10 "gold standard" WES and WGS datasets. Our results suggest that Bowtie2 performs significantly worse than other aligners and should not be used for medical variant calling. When other aligners were used, the accuracy of variant discovery mostly depended on the variant caller and not the read aligner. Among the tested variant callers, DeepVariant consistently showed the best performance and the highest robustness, with other state-of-the-art tools, i.e. Strelka2 and GATK with 1D convolutional neural network variant scoring, also showing high performance on both WES and WGS data. The results show surprisingly large differences in the performance of cutting-edge tools even on high confidence regions of the coding genome. This highlights the importance of regular benchmarking of quickly evolving tools and pipelines. Finally, we discuss the need for a more diverse set of gold standard genomes (e.g. of African, Hispanic, or mixed ancestry) that would allow to control for deep learning model overfitting. For similar reasons there is a need for better variant caller assessment in the repetitive regions of the coding genome.

bioinformatics↗

bioRxiv2020

Harnessing population-specific protein truncating variants to improve the annotation of loss-of-function alleles

Accurate annotation of putative loss-of-function (pLoF) variants is an important problem in human genomics and disease, which recently drew substantial attention. Since such variants in disease-related genes are under strong negative selection, their frequency across major ancestral groups is expected to be highly similar. In this study, we tested this assumption by systematically assessing the presence of highly population-specific protein-truncating variants (PTVs) in human genes using latest population-scale data. We discovered an unexpectedly high incidence of population-specific PTVs in all major ancestral groups. This does not conform to a recently proposed model, indicating either systemic differences in disease penetrance in different human populations, or a failure of current annotation criteria to accurately predict the loss-of-function potential of PTVs. We show that low-confidence pLoF variants are enriched in genes with non-uniform PTV count distribution, and developed a computational tool called LoFfeR that can efficiently predict tolerated pLoF variants. To evaluate the performance of LoFfeR, we use a set of known pathogenic and benign PTVs from the ClinVar database, and show that LoFfeR allows for a more accurate annotation of low-confidence pLoF variants compared to existing methods. Notably, only 4.4% of protein-truncating gnomAD SNPs in canonical transcripts can be filtered out using a recommended threshold value of the recently proposed pext score, while up to 10.9% of such variants are filtered using LoFfeR with the same false positive rate. Hence, we believe that LoFfeR can be used for additional filtering of low-confidence pLoF variants in population genomics and medical genetics studies.

genomics↗

bioRxiv2018

Systematic dissection of biases in whole-exome and whole-genome sequencing reveals major determinants of coding sequence coverage

Next generation DNA sequencing technologies are rapidly transforming the world of human genomics. Advantages and diagnostic effectiveness of the two most widely used resequencing approaches, whole exome (WES) and whole genome (WGS) sequencing, are still frequently debated. In our study we developed a set of statistical tools to systematically assess coverage of CDS regions provided by several modern WES platforms, as well as PCR-free WGS. Using several novel metrics to characterize exon coverage in WES and WGS, we showed that some of the WES platforms achieve substantially less biased CDS coverage than others, with lower within- and between-interval variation and virtually absent GC-content bias. We discovered that, contrary to a common view, most of the coverage bias in WES stems from mappability limitations of short reads, as well as exome probe design. We identified the ~ 500 kb region of human exome that could not be effectively characterized using short read technology. We also showed that the overall power for SNP and indel discovery in CDS region is virtually indistinguishable for WGS and best WES platforms. Our results indicate that deep WES (100x) using least biased technologies provides similar effective coverage (97% of 10x q10+ bases) and CDS variant discovery to the standard 30x WGS, suggesting that WES remains an efficient alternative to WGS in many applications. Our work could serve as a guide for selection of an up-to-date resequencing approach in human genomic studies.

genomics↗