Accurate prediction of damaging missense variants is critically important for interpreting a genome sequence. Although many methods have been developed, their performance has been limited. Recent advances in machine learning and the availability of large-scale population genomic sequencing data provide new opportunities to considerably improve computational predictions. Here we describe the graphical missense variant pathogenicity predictor (gMVP), a new method based on graph attention neural networks. Its main component is a graph with nodes that capture predictive features of amino acids and edges weighted by co-evolution strength, enabling effective pooling of information from the local protein context and functionally correlated distal positions. Evaluation of deep mutational scan data shows that gMVP outperforms other published methods in identifying damaging variants in TP53, PTEN, BRCA1 and MSH2. Furthermore, it achieves the best separation of de novo missense variants in neurodevelopmental disorder cases from those in controls. Finally, the model supports transfer learning to optimize gain- and loss-of-function predictions in sodium and calcium channels. In summary, we demonstrate that gMVP can improve interpretation of missense variants in clinical testing and genetic studies.
This is a preview of subscription content, access via your institution
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$29.99 / 30 days
cancel any time
Subscribe to this journal
Receive 12 digital issues and online access to articles
$99.00 per year
only $8.25 per issue
Rent or buy this article
Get just this article for as long as you need it
Prices may be subject to local taxes which are calculated during checkout
Pre-computed gMVP scores for all possible missense variants in canonical transcripts on human hg38 can be downloaded from https://www.dropbox.com/s/nce1jhg3i7jw1hx/gMVP.2021-02-28.csv.gz?dl=0. The training data of the main model were downloaded from http://www.discovehrshare.com/downloads (DiscovEHR), http://www.hgmd.cf.ac.uk/ac/index.php (HGMD), https://www.uniprot.org/docs/humpvar (UniProt) and https://ftp.ncbi.nlm.nih.gov/pub/clinvar/vcf_GRCh37/ (ClinVar). Other datasets supporting the findings of this study are available in the paper and the Supplementary Information.
The codes for the model design and training and testing procedure are available on GitHub (https://github.com/ShenLab/gMVP/) and Zenodo81.
Boettcher, S. et al. A dominant-negative effect drives selection of TP53 missense mutations in myeloid malignancies. Science 365, 599–604 (2019).
Huang, K. L. et al. Pathogenic germline variants in 10,389 adult cancers. Cell 173, 355–370.e14 (2018).
Jin, S. C. et al. Contribution of rare inherited and de novo variants in 2,871 congenital heart disease probands. Nat. Genet. 49, 1593–1601 (2017).
Satterstrom, F. K. et al. Large-scale exome sequencing study implicates both developmental and functional changes in the neurobiology of autism. Cell 180, 568–584.e23 (2020).
Kaplanis, J. et al. Evidence for 28 genetic disorders discovered by combining healthcare and research data. Nature 586, 757–762 (2020).
Rehm, H. L., Berg, J. S. & Plon, S. E. ClinGen and ClinVar—enabling genomics in precision medicine. Hum. Mutat. 39, 1473–1475 (2018).
He, X. et al. Integrated model of de novo and inherited genetic variants yields greater power to identify risk genes. PLoS Genet. 9, e1003671 (2013).
Nguyen, H. T. et al. Integrated Bayesian analysis of rare exonic variants to identify risk genes for schizophrenia and neurodevelopmental disorders. Genome Med. 9, 114 (2017).
Adzhubei, I., Jordan, D. M. & Sunyaev, S.R. Predicting functional effect of human missense mutations using PolyPhen-2. Curr. Protoc. Hum. Genet. https://doi.org/10.1002/0471142905.hg0720s76 (2013).
Carter, H., Douville, C., Stenson, P. D., Cooper, D. N. & Karchin, R. Identifying Mendelian disease genes with the variant effect scoring tool. BMC Genom. 14, S3 (2013).
Kircher, M. et al. A general framework for estimating the relative pathogenicity of human genetic variants. Nat. Genet. 46, 310–315 (2014).
Ioannidis, N. M. et al. REVEL: an ensemble method for predicting the pathogenicity of rare missense variants. Am. J. Hum. Genet. 99, 877–885 (2016).
Dong, C. et al. Comparison and integration of deleteriousness prediction methods for nonsynonymous SNVs in whole exome sequencing studies. Hum. Mol. Genet. 24, 2125–2137 (2015).
Jagadeesh, K. A. et al. M-CAP eliminates a majority of variants of uncertain significance in clinical exomes at high sensitivity. Nat. Genet. 48, 1581–1586 (2016).
Ionita-Laza, I., McCallum, K., Xu, B. & Buxbaum, J. D. A spectral approach integrating functional genomic annotations for coding and noncoding variants. Nat. Genet. 48, 214–220 (2016).
Qi, H. et al. MVP predicts the pathogenicity of missense variants by deep learning. Nat. Commun. 12, 510 (2021).
Sundaram, L. et al. Predicting the clinical impact of human mutation with deep neural networks. Nat. Genet. 50, 1161 (2018).
Samocha, K.E. et al. Regional missense constraint improves variant deleteriousness prediction. Preprint at bioRxiv https://doi.org/10.1101/148353 (2017).
Havrilla, J. M., Pedersen, B. S., Layer, R. M. & Quinlan, A. R. A map of constrained coding regions in the human genome. Nat. Genet. 51, 88–95 (2019).
Davydov, E. V. et al. Identifying a high fraction of the human genome to be under selective constraint using GERP plus. PLoS Comput. Biol. 6, e1001025 (2010).
Iqbal, S. et al. Comprehensive characterization of amino acid positions in protein structures reveals molecular effect of missense variants. Proc. Natl Acad. Sci. USA 117, 28201–28211 (2020).
Hicks, M., Bartha, I., di Iulio, J., Venter, J. C. & Telenti, A. Functional characterization of 3D protein structures informed by human genetic diversity. Proc. Natl Acad. Sci. USA 116, 8960–8965 (2019).
Sivley, R. M., Dou, X. Y., Meiler, J., Bush, W. S. & Capra, J. A. Comprehensive analysis of constraint on the spatial distribution of missense variants in human protein structures. Am. J. Hum. Genet. 102, 415–426 (2018).
Hopf, T. A. et al. Mutation effects predicted from sequence co-variation. Nat. Biotechnol. 35, 128–135 (2017).
Liang, S., Mort, M., Stenson, P.D., Cooper, D.N. & Yu, H. PIVOTAL: prioritizing variants of uncertain significance with spatial genomic patterns in the 3D proteome. Preprint at bioRxiv https://doi.org/10.1101/2020.06.04.135103 (2021).
Chang, M. T. et al. Accelerating discovery of functional mutant alleles in cancer. Cancer Discov. 8, 174–183 (2018).
Jia, X. et al. Massively parallel functional testing of MSH2 missense variants conferring Lynch syndrome risk. Am. J. Hum. Genet. 108, 163–175 (2021).
Findlay, G. M. et al. Accurate classification of BRCA1 variants with saturation genome editing. Nature 562, 217–222 (2018).
Mighell, T. L., Evans-Dutson, S. & O’Roak, B. J. A saturation mutagenesis approach to understanding PTEN lipid phosphatase activity and genotype–phenotype relationships. Am. J. Hum. Genet. 102, 943–955 (2018).
Kotler, E. et al. A systematic p53 mutation library links differential functional impact to cancer mutation pattern and evolutionary conservation. Mol. Cell 71, 178–190.e8 (2018).
de Juan, D., Pazos, F. & Valencia, A. Emerging methods in protein co-evolution. Nat. Rev. Genet. 14, 249–261 (2013).
Morcos, F. et al. Direct-coupling analysis of residue coevolution captures native contacts across many protein families. Proc. Natl Acad. Sci. USA 108, E1293–E1301 (2011).
Vaswani, A. et al. Attention is all you need. In 31st Conference on Neural Information Processing Systems 5998–6008 (NeurIPS, 2017).
Veličković, P. et al. Graph attention networks. In 6th International Conference on Learning Representations (Univ. Cambridge, 2018).
Cho, K. et al. Learning phrase representations using RNN encoder–decoder for statistical machine translation. In Proc. 2014 Conference on Empirical Methods in Natural Language Processing (Association for Computational Linguistics, 2014).
Stenson, P. D. et al. Human gene mutation database (HGMD (R)): 2003 update. Hum. Mutat. 21, 577–581 (2003).
Landrum, M. J. et al. ClinVar: public archive of relationships among sequence variation and human phenotype. Nucl. Acids Res. 42, D980–D985 (2014).
Mottaz, A., David, F. P., Veuthey, A. L. & Yip, Y. L. Easy retrieval of single amino-acid polymorphisms and phenotype information using SwissVar. Bioinformatics 26, 851–852 (2010).
Kingma, D. P. & Ba, J. Adam: a method for stochastic optimization. In 2015 International Conference on Learning Representations (ICLR, 2015).
Abadi, M. et al. TensorFlow: large-scale machine learning on heterogeneous distributed systems. Preprint at https://arxiv.org/abs/1603.04467 (2016).
Alirezaie, N., Kernohan, K. D., Hartley, T., Majewski, J. & Hocking, T. D. ClinPred: prediction tool to identify disease-relevant nonsynonymous single-nucleotide variants. Am. J. Hum. Genet. 103, 474–483 (2018).
Feng, B. J. PERCH: a unified framework for disease gene prioritization. Hum. Mutat. 38, 243–251 (2017).
Dewey, F. E. et al. Distribution and clinical impact of functional variants in 50,726 whole-exome sequences from the DiscovEHR Study. Science 354, aaf6814 (2016).
Iossifov, I. et al. The contribution of de novo coding mutations to autism spectrum disorder. Nature 515, 216–221 (2014).
De Rubeis, S. et al. Synaptic, transcriptional and chromatin genes disrupted in autism. Nature 515, 209–215 (2014).
Zuk, O. et al. Searching for missing heritability: designing rare variant association studies. Proc. Natl Acad. Sci. USA 111, E455–E464 (2014).
Heyne, H. O. et al. Predicting functional effects of missense variants in voltage-gated sodium and calcium channels. Sci. Transl. Med. 12, eaay6848 (2020).
Karczewski, K. J. et al. The mutational constraint spectrum quantified from variation in 141,456 humans. Nature 581, 434–443 (2020).
Abrusán, G. & Marsh, J. A. Alpha helices are more robust to mutations than beta strands. PLoS Comput. Biol. 12, e1005242 (2016).
Gao, M., Zhou, H. & Skolnick, J. Insights into disease-associated mutations in the human proteome through protein structural analysis. Structure 23, 1362–1369 (2015).
Li, S.-C., Goto, N. K., Williams, K. A. & Deber, C. M. Alpha-helical, but not beta-sheet, propensity of proline is determined by peptide environment. Proc. Natl Acad. Sci. USA 93, 6676–6681 (1996).
Senior, A. W. et al. Improved protein structure prediction using potentials from deep learning. Nature 577, 706–710 (2020).
Yang, J. Y. et al. Improved protein structure prediction using predicted interresidue orientations. Proc. Natl Acad. Sci. USA 117, 1496–1503 (2020).
Wang, S., Sun, S., Li, Z., Zhang, R. & Xu, J. Accurate de novo prediction of protein contact map by ultra-deep learning model. PLoS Comput. Biol. 13, e1005324 (2017).
Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021).
Kumar, S., Clarke, D. & Gerstein, M. B. Leveraging protein dynamics to identify cancer mutational hotspots using 3D structures. Proc. Natl Acad. Sci. USA 116, 18962–18970 (2019).
Anishchenko, I., Ovchinnikov, S., Kamisetty, H. & Baker, D. Origins of coevolution between residues distant in protein 3D structures. Proc. Natl Acad. Sci. USA 114, 9122–9127 (2017).
Tunyasuvunakool, K. et al. Highly accurate protein structure prediction for the human proteome. Nature 596, 590–596 (2021).
Rao, R. et al. MSA transformer. In Proc. 38th International Conference on Machine Learning 8844–8856 (PMLR, 2021).
Rives, A. et al. Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences. Proc. Natl Acad. Sci. USA 118, e2016239118 (2021).
Rao, R., Meier, J., Sercu, T., Ovchinnikov, S. & Rives, A. Transformer protein language models are unsupervised structure learners. In 2015 International Conference on Learning Representations (ICLR, 2015).
Lal, D. et al. Gene family information facilitates variant interpretation and identification of disease-associated genes in neurodevelopmental disorders. Genome Med. 12, 28 (2020).
Zhang, X. et al. Disease-specific variant pathogenicity prediction significantly improves variant interpretation in inherited cardiac conditions. Genet. Med. 23, 69–79 (2021).
Starita, L. M. et al. Variant interpretation: functional assays to the rescue. Am. J. Human Genet. 101, 315–325 (2017).
Brnich, S. E. et al. Recommendations for application of the functional evidence PS3/BS3 criterion using the ACMG/AMP sequence variant interpretation framework. Genome Med. 12, 3 (2019).
Hartl, D. L. & Clark, A. G. Principles of Population Genetics 4th edn (Sinauer Associates, 1989).
Cassa, C. A. et al. Estimating the selective effects of heterozygous protein-truncating variants from human exome data. Nat. Genet. 49, 806–810 (2017).
Charlesworth, B. & Hill, W. G. Selective effects of heterozygous protein-truncating variants. Nat. Genet. 51, 2 (2019).
Bycroft, C. et al. The UK Biobank resource with deep phenotyping and genomic data. Nature 562, 203–209 (2018).
Mulder, N. et al. H3Africa: current perspectives. Pharmgenomics Pers. Med. 11, 59–66 (2018).
Siepel, A. et al. Evolutionarily conserved elements in vertebrate, insect, worm, and yeast genomes. Genome Res. 15, 1034–1050 (2005).
Glorot, X., Bordes, A. & Bengio, Y. Deep sparse rectifier neural networks. In Proc. 14th International Conference on Artificial Intelligence and Statistics 315–323 (JMLR, 2011).
Ke, G., He, D. & Liu, T.-Y. Rethinking positional encoding in language pre-training. In 2021 International Conference on Learning Representations (ICLR, 2021).
Bateman, A. Uniprot: a universal hub of protein knowledge. Protein Sci. 28, 32–32 (2019).
Remmert, M., Biegert, A., Hauser, A. & Soding, J. HHblits: lightning-fast iterative protein sequence searching by HMM-HMM alignment. Nat. Methods 9, 173–175 (2012).
Herrero, J. et al. Ensembl comparative genomics resources. Database 2016, bav096 (2016).
Klausen, M. S. et al. NetSurfP-2.0: improved prediction of protein structural features by integrated deep learning. Proteins 87, 520–527 (2019).
Armean, I. M. et al. Enhanced access to extensive phenotype and disease annotation of genes and genetic variation in Ensembl. Eur. J. Human Genet. 27, 1721–1721 (2019).
McLaren, W. et al. The Ensembl variant effect predictor. Genome Biol. 17, 122 (2016).
Ge, R., Kakade, S. M., Kidambi, R. & Netrapalli, P. Rethinking learning rate schedules for stochastic optimization. In 2019 International Conference on Learning Representations (ICLR, 2018).
Zhang, H. & Shen, Y. ShenLab/gMVP: v1.0.0-alpha. Zenodo https://doi.org/10.5281/zenodo.7134878 (2022).
This work was supported by NIH grants (nos. R01GM120609, R03HL147197, U01HG008680 and K99HG011490) and the Columbia University Precision Medicine Joint Pilot Grants Program. We thank Y. Zhao, G. Zhong, M. AlQuraishi and D. Knowles for helpful discussions.
The authors declare no competing interests.
Peer review information
Nature Machine Intelligence thanks Xiaoming Liu, Wim Vranken, Amit R Majithia and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Test, Figs. 1–15 and Tables 11–18.
Supplementary Tables 1–10.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Zhang, H., Xu, M.S., Fan, X. et al. Predicting functional effect of missense variants using graph attention neural networks. Nat Mach Intell 4, 1017–1028 (2022). https://doi.org/10.1038/s42256-022-00561-w