DeepSVP: integration of genotype and phenotype for structural variant prioritization using deep learning

Abstract

Motivation Structural genomic variants account for much of human variability and are involved in several diseases. Structural variants are complex and may affect coding regions of multiple genes, or affect the functions of genomic regions in different ways from single nucleotide variants. Interpreting the phenotypic consequences of structural variants relies on information about gene functions, haploinsufficiency or triplosensitivity, and other genomic features. Phenotype-based methods to identifying variants that are involved in genetic diseases combine molecular features with prior knowledge about the phenotypic consequences of altering gene functions. While phenotype-based methods have been applied successfully to single nucleotide variants as well as short insertions and deletions, the complexity of structural variants makes it more challenging to link them to phenotypes. Furthermore, structural variants can affect a large number of coding regions, and phenotype information may not be available for all of them. Results We developed DeepSVP, a computational method to prioritize structural variants involved in genetic diseases by combining genomic and gene functions information. We incorporate phenotypes linked to genes, functions of gene products, gene expression in individual celltypes, and anatomical sites of expression, and systematically relate them to their phenotypic consequences through ontologies and machine learning. DeepSVP significantly improves the success rate of finding causative variants in several benchmarks and can identify novel pathogenic structural variants in consanguineous families. Availability https://github.com/bio-ontology-research-group/DeepSVP

Advancements in Genomic and Behavioral Neuroscience Analysis for the Study of Normal and Pathological Brain Function

“…Geneticists are now using deep learning approaches to efficiently identify quantitative trait loci and single nucleotide polymorphisms for thousands of emergent phenotypes, both physiological and behavioral. For example, DeepSVP is a deep learning algorithm that leverages genomic information obtained in ontologies and phenotypes obtained in mouse loss-of-function studies to accurately predict whether a structural variant in the human genome will be pathogenic ( Althagafi et al, 2021 ). Table 5 summarizes different, novel deep learning methods that are being used to characterize phenotypes based on genetics and morphology, in addition to behavior.…”

RExPRT: a machine learning tool to predict pathogenicity of tandem repeat loci

“…Another tool, SVpath, predicts the pathogenicity of exonic SVs by incorporating features that are based on functional impact scores of overlapping SNVs, as well as gene level and transcriptomics scores [ 15 ]. Developing this approach further, DeepSVP integrates ranking of noncoding variants and incorporates phenotype information using a deep learning approach to improve the selection of patient-specific variants in a more precise manner [ 16 ].…”

RExPRT: a machine learning tool to predict pathogenicity of tandem repeat loci

“…15 Developing this approach further, DeepSVP integrates ranking of noncoding variants and incorporates phenotype information using a deep learning approach to improve the selection of patient-specific variants in a more precise manner. 16 In the field of TRs, efforts to prioritize variants have focused on an underlying assumption that TR constraint correlates with pathogenicity. Gymrek et al showed this to be true for select early-onset disease loci such as RUNX2 and HOXD13.…”

“…15 Developing this approach further, DeepSVP integrates ranking of noncoding variants and incorporates phenotype information using a deep learning approach to improve the selection of patient-specific variants in a more precise manner. 16…”