Disease-related mutations predicted to impact protein function

BMC Genomics

Proceedings

Yana Bromberg¹  Dominik Achten²  Christian Schaefer³  Burkhard Rost⁴ 

[1] Department of Biochemistry and Microbiology, School of Environmental and Biological Sciences, Rutgers University, 08901, New Brunswick, NJ, USA;TUM, Bioinformatics - i12, Informatics, Boltzmannstrasse 3, 85748, Garching/Munich, Germany;TUM, Bioinformatics - i12, Informatics, Boltzmannstrasse 3, 85748, Garching/Munich, Germany;TUM Graduate School of Information Science in Health (GSISH), Boltzmannstr. 11, 85748, Garching/Munich, Germany;TUM, Bioinformatics - i12, Informatics, Boltzmannstrasse 3, 85748, Garching/Munich, Germany;TUM Graduate School of Information Science in Health (GSISH), Boltzmannstr. 11, 85748, Garching/Munich, Germany;Institute of Advanced Study (TUM-IAS), Lichtenbergstr. 2a, 85748, Garching/Munich, Germany;Columbia University, Department of Biochemistry and Molecular Biophysics & New York Consortium on Membrane Protein Structure (NYCOMPS), 701 West, 168th Street, 10032, New York, NY, USA;

关键词: Functional Effect;    Decision Boundary;    Disease Association;    Functional Impact;    Impact Function;   

DOI  :  10.1186/1471-2164-13-S4-S11

来源: Springer

【 摘 要 】

BackgroundNon-synonymous single nucleotide polymorphisms (nsSNPs) alter the protein sequence and can cause disease. The impact has been described by reliable experiments for relatively few mutations. Here, we study predictions for functional impact of disease-annotated mutations from OMIM, PMD and Swiss-Prot and of variants not linked to disease.ResultsMost disease-causing mutations were predicted to impact protein function. More surprisingly, the raw predictions scores for disease-causing mutations were higher than the scores for the function-altering data set originally used for developing the prediction method (here SNAP). We might expect that diseases are caused by change-of-function mutations. However, it is surprising how well prediction methods developed for different purposes identify this link. Conversely, our predictions suggest that the set of nsSNPs not currently linked to diseases contains very few strong disease associations to be discovered.ConclusionsFirstly, annotations of disease-causing nsSNPs are on average so reliable that they can be used as proxies for functional impact. Secondly, disease-causing nsSNPs can be identified very well by methods that predict the impact of mutations on protein function. This implies that the existing prediction methods provide a very good means of choosing a set of suspect SNPs relevant for disease.

【 授权许可】

CC BY   
© Schaefer et al; licensee BioMed Central Ltd. 2012