A method and server for predicting damaging missense mutations
By Ivan A. Adzhubei, S. Schmidt, Leonid Peshkin, V. Ramensky, A. Gerasimova, Peer Bork, Alexey S. Kondrashov, Shamil R. Sunyaev. 2010
13,139 citations recorded by the source corpus.
To the Editor: Applications of rapidly advancing sequencing technologies exacerbate the need to interpret individual sequence variants. Sequencing of phenotyped clinical subjects will soon become a method of choice in studies of the genetic causes of Mendelian and complex diseases. New exon capture techniques will direct sequencing efforts towards the most informative and easily interpretable protein-coding fraction of the genome. Thus, the demand for computational predictions of the impact of protein sequence variants will continue to grow. Here we present a new method and the corresponding software tool, PolyPhen-2 (http://genetics.bwh.harvard.edu/pph2/), which is different from the early tool PolyPhen1 in the set of predictive features, alignment pipeline, and the method of classification (Fig. 1a). PolyPhen-2 uses eight sequence-based and three structure-based predictive features (Supplementary Table 1) which were selected automatically by an iterative greedy algorithm (Supplementary Methods). Majority of these features involve comparison of a property of the wild-type (ancestral, normal) allele and the corresponding property of the mutant (derived, disease-causing) allele, which together define an amino acid replacement. Most informative features characterize how well the two human alleles fit into the pattern of amino acid replacements within the multiple sequence alignment of homologous proteins, how distant the protein harboring the first deviation from the human wild-type allele is from the human protein, and whether the mutant allele originated at a hypermutable site2. The alignment pipeline selects the set of homologous sequences for the analysis using a clustering algorithm and then constructs and refines their multiple alignment (Supplementary Fig. 1). The functional significance of an allele replacement is predicted from its individual features (Supplementary Figs. 2–4) by Naive Bayes classifier (Supplementary Methods). Figure 1 PolyPhen-2 pipeline and prediction accuracy. (a) Overview of the algorithm. (b) Receiver operating characteristic (ROC) curves for predictions made by PolyPhen-2 using five-fold cross-validation on HumDiv (red) and HumVar3 (light green). UniRef100 (solid ... We used two pairs of datasets to train and test PolyPhen-2. We compiled the first pair, HumDiv, from all 3,155 damaging alleles with known effects on the molecular function causing human Mendelian diseases, present in the UniProt database, together with 6,321 differences between human proteins and their closely related mammalian homologs, assumed to be non-damaging (Supplementary Methods). The second pair, HumVar3, consists of all the 13,032 human disease-causing mutations from UniProt, together with 8,946 human nsSNPs without annotated involvement in disease, which were treated as non-damaging. We found that PolyPhen-2 performance, as presented by its receiver operating characteristic curves, was consistently superior compared to PolyPhen (Fig. 1b) and it also compared favorably with the three other popular prediction tools4–6 (Fig. 1c). For a false positive rate of 20%, PolyPhen-2 achieves the rate of true positive predictions of 92% and 73% on HumDiv and HumVar, respectively (Supplementary Table 2). One reason for a lower accuracy of predictions on HumVar is that nsSNPs assumed to be non-damaging in HumVar contain a sizable fraction of mildly deleterious alleles. In contrast, most of amino acid replacements assumed non-damaging in HumDiv must be close to selective neutrality. Because alleles that are even mildly but unconditionally deleterious cannot be fixed in the evolving lineage, no method based on comparative sequence analysis is ideal for discriminating between drastically and mildly deleterious mutations, which are assigned to the opposite categories in HumVar. Another reason is that HumDiv uses an extra criterion to avoid possible erroneous annotations of damaging mutations. For a mutation, PolyPhen-2 calculates Naive Bayes posterior probability that this mutation is damaging and reports estimates of false positive (the chance that the mutation is classified as damaging when it is in fact non-damaging) and true positive (the chance that the mutation is classified as damaging when it is indeed damaging) rates. A mutation is also appraised qualitatively, as benign, possibly damaging, or probably damaging (Supplementary Methods). The user can choose between HumDiv- and HumVar-trained PolyPhen-2. Diagnostics of Mendelian diseases requires distinguishing mutations with drastic effects from all the remaining human variation, including abundant mildly deleterious alleles. Thus, HumVar-trained PolyPhen-2 should be used for this task. In contrast, HumDiv-trained PolyPhen-2 should be used for evaluating rare alleles at loci potentially involved in complex phenotypes, dense mapping of regions identified by genome-wide association studies, and analysis of natural selection from sequence data, where even mildly deleterious alleles must be treated as damaging.
Subfields: Genetics, Molecular Biology, Computational Biology & Bioinformatics
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Genetics
- Eric Boerwinkle
- B. Feng
- Paul D. Thomas
- Xiaoming Liu
- Mauno Vihinen
- Russ B. Altman
- Haiming Tang
- Yana Bromberg
In Molecular Biology
- Burkhard Rost
- Yang Zhang
- Mauno Vihinen
- Eric Boerwinkle
- Yana Bromberg
- Russ B. Altman
- Paul D. Thomas
- Andrej Sali
In Computational Biology & Bioinformatics
- Burkhard Rost
- Yana Bromberg
- Yang Zhang
- Michael Heinzinger
- Christopher Sander
- Debora Marks
- Russ B. Altman
- Martin Weigt
Citations in the corpus
- Modeling aspects of the language of life through transfer-learning protein sequences (2019, background): “…aspects of protein function improved by combining evolutionary information and machine learning, including predictions of subcellular localization (aka cellular compartment or CC in GO [21, 22]), protein interaction sites [23 – 25], and the effects of sequence variation upon function [26, 27].”
- CADD: predicting the deleteriousness of variants throughout the human genome (2018, positive): “Examples of annotations include transcript information like distance to exon-intron boundaries, DNase hypersensitivity, transcription factor binding, expression levels in commonly studied cell lines and amino acid substitution scores for protein coding sequences like Grantham (20), SIFT (21) and Pol”
- Quantitative missense variant effect prediction using large-scale mutagenesis data (2017, neutral): “Thus, variant effect predictors such as PolyPhen2 (Adzhubei et al., 2010), SIFT (Sim et al., 2012), SNAP2 (Hecht et al., 2015), Evolutionary Action (Katsonis and Lichtarge, 2014), CADD (Kircher et al., 2014), and a host of others (Tang and Thomas, 2016) will continue to be widely used to predict…”
- Coevolutionary Landscape Inference and the Context-Dependence of Mutations in Beta-Lactamase TEM-1 (2015, background): “Critically important residues tend to be conserved, while unfavorable residues are observed less frequently.”
- Better prediction of functional effects for sequence variants (2015, positive): “Methods have improved significantly by using more protein and variant annotations, as demonstrated in particular in the advance from PolyPhen [12] to PolyPhen-2 [13].”
- Predicting Functional Effect of Human Missense Mutations Using PolyPhen-2 (2013, methodology): “PolyPhen-2 (Adzhubei et al., 2010) is an automatic tool for prediction of the possible impact of an amino acid substitution on the structure and function of a human protein.”
- A general framework for estimating the relative pathogenicity of human genetic variants (2014, background): “Due to the overlap of ClinVar and ESP variants with the PolyPhen training data set, we trained a separate classifier without the PolyPhen features and we also checked the performance on the subset of ClinVar and ESP variants not used for PolyPhen training.”
- SIFT web server: predicting effects of amino acid substitutions on proteins (2012, methodology): “We assess its performance on the HumDiv and HumVar data sets, which were created by the authors of PolyPhen-2 (20).”
- Language models enable zero-shot prediction of the effects of mutations on protein function (2021, methodology): “Models have been designed specifically for proteins, using feature engineering such as Envision [36] and PolyPhen-2 [37], ensemble methods such as Revel [38], MPC [39], CADD [40], and M-CAP [41], language models such as UniRep [23, 16] and ESM [12], and other representation learning approaches [42, ”
- RAD51 and Breast Cancer Susceptibility: No Evidence for Rare Variant Association in the Breast Cancer Family Registry Study (2012, methodology): “1 software, using its precompiled alignment [57].”