Machine-learning of complex evolutionary signals improves classification of SNVs

Abstract

Conservation is a strong predictor for the pathogenicity of single-nucleotide variants (SNVs). However, some positions that present complex conservation patterns across vertebrates stray from this paradigm. Here, we analyzed the association between complex conservation patterns and the pathogenicity of SNVs in the 115 disease-genes that had sufficient variant data. We show that conservation is not a one-rule-fits-all solution since its accuracy highly depends on the analyzed set of species and genes. For example, pairwise comparisons between the human and 99 vertebrate species showed that species differ in their ability to predict the clinical outcomes of variants among different genes using conservation. Furthermore, certain genes were less amenable for conservation-based variant prediction, while others demonstrated species that optimize prediction. These insights led to developing EvoDiagnostics, which uses the conservation against each species as a feature within a random-forest machine-learning classification algorithm. EvoDiagnostics outperformed traditional conservation algorithms, deep-learning based methods and most ensemble tools in every prediction-task, highlighting the strength of optimizing conservation analysis per-species and per-gene. Overall, we suggest a new and a more biologically relevant approach for analyzing conservation, which improves prediction of variant pathogenicity.

Pathogenic Variants of Scavenger Receptor CD36 Lead to Decreased Efferocytosis and Predispose to Myocarditis Following Vaccination With Pfizer-BioNTech BNT162b2 Against Coronavirus Infection (COVID-19)

“…Conservation patterns are strong predictors of pathogenicity for single nucleotide variants. EvoDiagnostics 3 detects these patterns using a machine-learning algorithm and accurately classifies CD36 single nucleotide variants (area under the receiver operating characteristic curve=0.93). Gly217Arg and Asp184Asn were found to have conservation patterns similar to known pathogenic variants.…”

Cutting-Edge AI Technologies Meet Precision Medicine to Improve Cancer Care

“…Sequence-based techniques [ 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 ] are used to construct the most common tools for predicting the pathogenicity of genetic variations. For instance, REVEL [ 10 ] employs a random forest method based on ensemble methods with 13 pathogenicity predictors.…”

“…CADD [ 12 ] is another ensemble method that integrates several scoring algorithms using a linear kernel support vector machine. VEST4 [ 8 ], EvoDiagnostics [ 18 ], and MetaSVM [ 9 ] are well-known random forest (RF) and support vector machine (SVM) prediction tools. Random forests and SVM may handle linear and non-linear data.…”

Using evolutionary data to make sense of macromolecules with a “face‐lifted” ConSurf

“…Exploiting evolutionary data to detect functional regions in proteins and in nucleic acids is very commonly used (Capra et al 2009 ; del Sol Mesa et al 2003 ; Gallet et al 2000 ; Landgraf et al 2001 ; Lichtarge et al 1996a ; Lichtarge et al 1996b ; Lichtarge et al 1997 ; Valdar 2002 ). Evolutionary rates are often used in genomics analyses to predict the pathogenicity of single‐nucleotide variants identified in patient samples (Labes et al 2022 and references therein). They can also be used in protein engineering efforts (Pavelka et al 2009 ).…”