Assessment of computational methods for predicting the effects of missense mutations in human cancers

Abstract

Background Recent advances in sequencing technologies have greatly increased the identification of mutations in cancer genomes. However, it remains a significant challenge to identify cancer-driving mutations, since most observed missense changes are neutral passenger mutations. Various computational methods have been developed to predict the effects of amino acid substitutions on protein function and classify mutations as deleterious or benign. These include approaches that rely on evolutionary conservation, structural constraints, or physicochemical attributes of amino acid substitutions. Here we review existing methods and further examine eight tools: SIFT, PolyPhen2, Condel, CHASM, mCluster, logRE, SNAP, and MutationAssessor, with respect to their coverage, accuracy, availability and dependence on other tools. Results Single nucleotide polymorphisms with high minor allele frequencies were used as a negative (neutral) set for testing, and recurrent mutations from the COSMIC database as well as novel recurrent somatic mutations identified in very recent cancer studies were used as positive (non-neutral) sets. Conservation-based methods generally had moderately high accuracy in distinguishing neutral from deleterious mutations, whereas the performance of machine learning based predictors with comprehensive feature spaces varied between assessments using different positive sets. MutationAssessor consistently provided the highest accuracies. For certain combinations metapredictors slightly improved the performance of included individual methods, but did not outperform MutationAssessor as stand-alone tool. Conclusions Our independent assessment of existing tools reveals various performance disparities. Cancer-trained methods did not improve upon more general predictors. No method or combination of methods exceeds 81% accuracy, indicating there is still significant room for improvement for driver mutation prediction, and perhaps more sophisticated feature integration is needed to develop a more robust tool.

Characterizing and comparing missense variants in monogenic disease and in cancer

“…An additional four monogenic disease missense analysis methods and three developed specifically for cancer missense analysis show the same pattern (Table 1). Previous studies have also shown similar results for cancer data [62,63]. Figure 1B shows that in contrast to the results for the disease mutations, interspecies variants in these three sets of genes have a uniformly low predicted deleterious fraction (0.03 -0.10), with no significant differences between monogenic disease and cancer.…”

ConsensusDriver improves upon individual algorithms for predicting driver alterations in different cancer types and individual patients – a toolbox for precision oncology

“…In terms of driver prediction methods, we attempted to be as comprehensive as possible, though some methods had to be excluded due to feasibility issues (see Methods). In total, we evaluated 18 methods that could be used for driver prediction (Figure 1b), classifying these methods into (i) methods that belong to the Functional Impact (FI) category (primarily designed to identify function altering mutations but have been used for predicting drivers [29,30,46]) such as SIFT [22], PolyPhen2 (PP2) [23], MutationTaster (MT) [24] and MutationAssessor (MA) [25], (ii) methods that tailor this idea to cancer by learning specific models (Functional Impact in Cancer; FIC) such as CHASM [26], transFIC (TF) [27] and fathmm (FH) [28], (iii) methods that use cohort based analysis to detect signals of positive selection (CBA) such as ActiveDriver (AD) [36], MutSigCV (MCV) [31], MuSiC (MUS) [32], OncodriveCLUST (OCL) [33] and OncodriveFM (OF) [34] (all point mutation based), (iv) methods that integrate mutation data with transcriptomic data by looking for mutation-expression correlations (MEC) such as Conexic (CON) [38], OncodriveCIS (OCI) [39] and S2N [40], and finally (v) methods that use information from gene/protein interaction networks to analyze the effect of mutations such as NetBox (NB) [44], HotNet2 (HN2) [45], DriverNet (DN) [8], DawnRank (DR) [9] and OncoIMPACT (OI) [10]. We evaluated these 18 methods in predicting cancer drivers in patient cohorts and in individual patients.…”

“…These methods predict functional/driver mutations in each sample independently and their relative strengths have been studied in previous work [29,30]. With the availability of large and heterogeneous cancer genomic datasets, newer methods have focused on cohort based analysis to search for biases in mutation frequency indicative of positive selection in driver genes (CBA) [31][32][33][34][35][36] (compared in [37]), or mined for mutation-expression correlations to highlight driver CNAs (MEC) [38][39][40] (jointly evaluated in [41]).…”

Predicting Protein Thermostability Upon Mutation Using Molecular Dynamics Timeseries Data

“…A number of approaches have been utilized to predict protein stability upon mutation. Some of most widely utilized computational techniques employ protein sequence data [14]. Both conserved and non-conserved regions exist in protein sequences across multiple organisms.…”