PDF《Analysis and prediction of protein》

Analysis and prediction of protein folding energy changes upon mutation by element speci?c persistent homology Zixuan Cang1 , and Guo-Wei Wei1,2,3 ?

1 Department of Mathematics Michigan State University, MI 48824, USA

2 Department of Biochemistry and Molecular Biology Michigan State University, MI 48824, USA

3 Department of Electrical and Computer Engineering Michigan State University, MI 48824, USA Abstract Motivation: Site directed mutagenesis is widely used to understand the structure and function of biomolecules.

Computational prediction of protein mutation impacts offers a fast, economical and potentially accurate alterna- tive to laboratory mutagenesis. Most existing methods rely on geometric descriptions, this work introduces a topology based approach to provide an entirely new representation of protein mutation impacts that could not be obtained from conventional techniques. Results: Topology based mutation predictor (T-MP) is introduced to dramatically reduce the geometric com- plexity and number of degrees of freedom of proteins, while element speci?c persistent homology is proposed to retain essential biological information. The present approach is found to outperform other existing methods in globular protein mutation impact predictions. A Pearson correlation coef?cient of 0.82 with an RMSE of 0.92 kcal/mol is obtained on a test set of

350 mutation samples. For the prediction of membrane protein stability changes upon mutation, the proposed topological approach has a 84% higher Pearson correlation coef?cient than the current state-of-the-art empirical methods, achieving a Pearson correlation of 0.57 and an RMSE of 1.09 kcal/mol in a 5-fold cross validation on a set of

223 membrane protein mutation samples. I Introduction Mutagenesis, as a basic biological process that changes the genetic information of organisms, serves as a primary source for many kinds of cancer and heritable diseases, as well as a driving force for natural evolu- tion.1C3 For example, more than

60 human hereditary diseases are directly related to mutagenesis in proteases and their natural inhibitors.4 Additionally, mutagenesis often leads to drug resistance.5 Mutation, as a result of mutagenesis, can either occur spontaneously in nature or be caused by the exposure to a large dose of muta- gens in living organisms. In laboratories, site directed mutagenesis analysis is a vital experimental procedure for exploring protein functional changes in enzymatic catalyzing, structural supporting, ligand binding, and sig- naling.6 Nonetheless, site directed mutagenesis analysis is both time-consuming and expensive. Additionally, site directed mutagenesis measurements for one speci?c mutation obtained from different approaches may vary dramatically, particularly for membrane protein mutations. Computational prediction of protein mutation impacts is an important alternative to experimental mutagen- esis analysis for the systematical exploration of protein structural instabilities, functions, disease connections, and organism evolution pathways.7 A major advantage of these approaches is that they provide an economical, fast, and potentially accurate alternative to site directed mutagenesis experiments. Many state-of-the-art meth- ods have been developed in the past decade, including I-Mutant,8 PoPMuSiC,9 knowledge-modi?ed MM/PBSA approach,10 Rosetta (high) protocols,11 FoldX (3.0, beta 6.1),7 SDM,12 DUET,13 PPSC (Prediction of Protein Stability, version 1.0) with the

8 (M8) and

47 (M47) feature sets,14 PROVEAN,15 ELASPIC,16 STRUM,17 and EASE-MM.18 In general, computational approaches can be classi?ed into three major classes. Among them, physics based methods typically make use of molecular mechanics (MM), quantum mechanics (QM), or multi- scale implicit solvent models and QM/MM approaches. These approaches elucidate the fundamental of physics ?Address correspondences to Guo-Wei Wei. E-mail:[email protected]