VarCoPP2.0#

Via the oligopipe predict submodule

VarCoPP stands for Variant Combination Pathogenicity Predictor. It is a machine-learning method that predicts the pathogenicity of any bi-locus variant combination (i.e. a combination of two to four variant alleles between two genes).

The method has been published in the PNAS journal: https://doi.org/10.1073/pnas.1815601116. The first version of the model was later improved to VarCoPP2.0 using new training data, new features and a different type of model. VarCoPP2.0 has been published in BMC Bioinformatics: https://doi.org/10.1186/s12859-023-05291-3.

Based on VarCoPP, a bi-locus variant combination can either be candidate disease-causing or neutral.

VarCoPP has been trained separately for the GRCh37/hg19 and GRCh38/hg38 genome assembly, the correct model will run based on your assembly choice in the input.

NOTE: The use of VarCoPP for the analysis of complete patient exomes is NOT recommended, as the method is not fine-tuned for exomes. Please restrict your analysis to relevant gene panels for the disease of interest. If your VCF contains the complete exome of an individual, you can pass it to oligopipe predict as it is, and use the Variant Filtering and Gene Filtering options.

Alternatively, you can run a HOP prioritization analysis on a complete exome using oligopipe prioritize (see HOP).

Structure of VarCoPP#

ALGORITHM

VarCoPP2.0 is a Balanced Random Forest (RF) predictor that consists of 400 decision trees.

TRAINING DATA

The VarCoPP2.0 predictor has been trained on the pathogenic variant combinations present in the Oligogenic Diseases Database (OLIDA) against a large subset of variant data derived from control individuals of the 1000 Genomes Project (1KGP).

The variant types that were used for training were the same for both OLIDA and 1KGP: exonic and splicing variants of up to 3.5% MAF, while all genes were protein coding genes.

RESULT CALCULATION

VarCoPP2.0 outputs a VarCoPP score, i.e. the probability (value between 0 and 1) that a variant combination belongs to the disease-causing class. If this probability is above 0.50 (hg19) or 0.4575 (hg38), the model predicts that the combination is disease-causing.

Prediction features#

VarCoPP uses different variant, gene and gene pair biological features to make the predictions.

Feature	Source	Feature abbreviation	Gene / Variant allele
CADD raw score PMID: 24487276	CADD v1.6	CADD1 CADD2 CADD3 CADD4	Gene A / Variant allele 1 Gene A / Variant allele 2 Gene B / Variant allele 1 Gene B / Variant allele 2
Gene haploinsufficiency prediction PMID: 28137713	dbNSFP v4.1	HIPred_A HIPred_B	Gene A Gene B
Inheritance specific pathogenicity prediction for autosomal dominant (AD), autosomal recessive (AR), and X-linked (XL) modes of inheritance PMID: 27354691	publication supplementary data	ISPP_AD_A ISPP_AD_B ISPP_AR_A ISPP_AR_B ISPP_XL_A	Gene A Gene B Gene A Gene B Gene A
Selective pressure (dN/dS) PMID: 26896847	Ensembl v99	dN_dS_A	Gene A
Biological distance PMID: 24694260	Human Gene Connectome v12.2015	Biol_Dist	Gene pair AB
Biological Process similarity PMID: 10802651, 18460186	Gene Ontology v2021-12	BP_sim	Gene pair AB
Distance in an in-house developed Knowledge Graph PMID: 37644440	BOCK v1.0	KG_dist	Gene pair AB

Confidence zones#

With VarCoPP2.0 we have defined 99% and 99.9%confidence zones, delimited by the prediction threshold, which provide a probability of whether a particular combination predicted as candidate disease-causing, is actually a True Positive (TP) result. This indication can be useful for further evaluation and filtering of the predictions.

These confidence zones were created by testing neutral bi-locus combinations from the 1000 Genomes Project and obtaining the minimal prediction probability that gave 1% and 0.1% False Positives respectively. If a combination falls into either one of the two zones, a coloured indication will appear in the summary results.

99%-confidence zone

Requires VarCoPP score ≥ 0.743 (hg19) or ≥ 0.647 (hg38). If a digenic combination falls inside this zone, it has 99% probability of being a TP result.

99.9%-confidence zone

Requires VarCoPP score ≥ 0.891 (hg19) or ≥ 0.85 (hg38). If a digenic combination falls inside this zone, it has 99.9% probability of being a TP result.