X-CNV: genome-wide prediction of the pathogenicity of copy number variations

Abstract

Background Gene copy number variations (CNVs) contribute to genetic diversity and disease prevalence across populations. Substantial efforts have been made to decipher the relationship between CNVs and pathogenesis but with limited success. Results We have developed a novel computational framework X-CNV (www.unimd.org/XCNV), to predict the pathogenicity of CNVs by integrating more than 30 informative features such as allele frequency (AF), CNV length, CNV type, and some deleterious scores. Notably, over 14 million CNVs across various ethnic groups, covering nearly 93% of the human genome, were unified to calculate the AF. X-CNV, which yielded area under curve (AUC) values of 0.96 and 0.94 in training and validation sets, was demonstrated to outperform other available tools in terms of CNV pathogenicity prediction. A meta-voting prediction (MVP) score was developed to quantitively measure the pathogenic effect, which is based on the probabilistic value generated from the XGBoost algorithm. The proposed MVP score demonstrated a high discriminative power in determining pathogenetic CNVs for inherited traits/diseases in different ethnic groups. Conclusions The ability of the X-CNV framework to quantitatively prioritize functional, deleterious, and disease-causing CNV on a genome-wide basis outperformed current CNV-annotation tools and will have broad utility in population genetics, disease-association studies, and diagnostic screening.

StrVCTVRE: A supervised learning method to predict the pathogenicity of human genome structural variants

“…The poor performance of transcript consequence from VEP reinforces the known limitations of prioritizing variants with sequence ontology terms in isolation. We also evaluated X-CNV 35 and CADD-SV 36 on this test dataset ( Figure S6 ). The AUC for X-CNV was 0.68, and the AUC for CADD-SV was 0.70.…”

“…AnnotSV was run with human annotation and default settings. We retrieved X-CNV 35 on September 27, 2021, and it was run with default settings on variants converted to GRCh37 via the UCSC liftover tool. We retrieved CADD-SV 36 v1.0 on September 13, 2021, and it was run with default settings.…”

SvAnna: efficient and accurate pathogenicity prediction of coding and regulatory structural variants in long-read genome sequencing

“…We ran the benchmark on a case basis; we added all variants of the case report (i.e., both SVs in case of compound heterozygous genotype) in turn to one of ten background VCF files, and we recorded the median SV rank from the ten VCF files. We used the benchmarking schema to compare SvAnna with AnnotSV [ 27 ], X-CNV [ 28 ], SvScore [ 29 ], and ClassifyCNV [ 30 ]. The benchmark was performed using the pipeline engine Nextflow [ 31 ].…”

“…X-CNV [ 28 ] uses extreme gradient boosted trees to predict deleteriousness of deletions and duplications in the form of meta-voting prediction (MVP) score. The MVP score integrates several dozens of features, including SV characteristics (type, allele frequency based on DGV, dbVar), functional deleteriousness predictions, non-coding features (CTCF, cCREs, pELS, and dELS elements), and genome-wide annotations.…”

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

“…Methods that evaluate the functional consequence of SVs in individual genomes use different strategies. Several approaches include genomic information, such as variant length, haploinsufficiency measures or GC contents, to separate pathogenic from benign SVs ( Hehir-Kwa et al , 2010 ; Kleinert and Kircher, 2021 ; Sharo et al , 2020 ; Zhang et al , 2021 ). Furthermore, the predicted pathogenicity of deleterious single nucleotide variants within an SV can be used to estimate pathogenicity of SVs ( Ganel et al , 2017 ).…”