Comprehensive characterization of amino acid positions in protein structures reveals molecular effect of missense variants

• DOI: 10.1073/pnas.2002660117
• Publication Date: 2020-10-27T00:42:50Z
• Authors: Sumaiya Iqbal; Eduardo Pérez-Palma; Jakob B. Jespersen; Patrick May; David Hoksza; Henrike O. Heyne; Shehab S. Ahmed; Zaara T. Rifat; M. Sohel Rahman; Kasper Lage; Aarno Palotie; Jeffrey R. Cottrell; Florence F. Wagner; Mark J. Daly; Arthur J. Campbell; Dennis Lal
• Keywords: Models, Molecular; PREDICTION; Protein Conformation; sequence variation; Mutation, Missense; Protein function; SEQUENCE; Gene; DISEASE; Machine Learning; Computational biology; 03 medical and health sciences; Biochemistry, Genetics and Molecular Biology; Genetics; Humans; Missense mutation; Amino Acid Sequence; RNA Sequencing Data Analysis; Ribosome Structure and Translation Mechanisms; VERSION; Molecular Biology; Biology; Biochemistry, cell and molecular biology; SITES; 0303 health sciences; MUTATIONS; BRCA1 Protein; 3D mutational hotspot; PTEN Phosphohydrolase; Computational Biology; Proteins; Life Sciences; Biological Sciences; Standards and Guidelines for Genetic Variant Interpretation; REGIONS; EVOLUTION; Functional Genomics; Biomedicine; machine learning; Phenotype; Genetics, developmental biology, physiology; FOS: Biological sciences; protein structure and function; missense variant interpretation; disease variation effect; PATHOGENICITY

Significance Recent large-scale sequencing efforts have enabled the detection of millions missense variants. Elucidating their functional effect is crucial importance but challenging. We approach this problem by performing a wide-scale characterization variants from 1,330 disease-associated genes using >14,000 protein structures. identify 3D features associated with pathogenic and benign that unveiled mutations’ at molecular level. further extend our analysis to account for different essential structural regions in proteins functions. By analyzing 24 gene groups encoding families, we capture function-specific characteristics variants, which match experimental readouts. show results derived data will effectively inform variant interpretation.