Abstract—Benefiting from recent advancements of the next generation sequencing technology, it becomes more and more feasible to directly sequence candidate genetic regions and even the whole genome to get the information about rare genetic variants. Although several statistical methods have been developed to identify potential associations between multiple rare variants and a given disease of interest, these methods are quite sensitive to the inclusion of non-functional variants in their statistical analysis. In order to enhance the performance of these statistical methods for uncovering disease-associated rare variants, it is suggested that bioinformatics tools or filters should be adopted to make functional predictions of the variants before statistical analysis. In this paper, we propose to prioritize candidate genetic variants according to the guilt-by-association principle, which depends on the assumption that genetic variants associated with the same disease share some common physiochemical properties. Focusing on a specific type of genetic variants called single amino acid polymorphisms (SAAPs), we take advantages of 8 similarity measures based on physiochemical features of amino acids, sequence information of proteins, and multiple sequence alignment of protein families to illustrate the power of prioritizing candidate SAAPs for specific diseases. Systematic validation experiments demonstrate that our proposed approach is competent for effectively detecting associations between SAAPs and query diseases, while using the Canberra distance to measure the similarity between SAAPs can achieve the highest performance among all the methods compared.

Index Terms—Guilt-by-Association,Similarity, Single Amino Acid Polymorphisms (SAAPs), Prioritization.

Jiaxin Wu is with MOE Key laboratory of Bioinformatics and Bioinformatics Division, TNLIST/Department of Automation, Tsinghua University, Beijing 100084, China.
Mingxin Gan is with School of Economics and Management, University of Science and Technology Beijing, Beijing 100083, China.
Wangshu Zhang is with MOE Key laboratory of Bioinformatics and Bioinformatics Division, TNLIST/Department of Automation, Tsinghua University, Beijing 100084, China.
Rui Jiang is with MOE Key laboratory of Bioinformatics and Bioinformatics Division, TNLIST/Department of Automation, Tsinghua University, Beijing 100084, China. (FIT 1-107, Tsinghua University, Beijing 100084, China. e-mail: ruijiang@tsinghua.edu.cn).
Rui Jiang: To whom correspondence should be addressed. (FIT 1-107, Tsinghua University, Beijing 100084, China. E-mail: ruijiang@tsinghua.edu.cn)

Cite: Jiaxin Wu, Mingxin Gan, Wangshu Zhang, and Rui Jiang, "Prediction of Disease-associated Single Amino Acid Polymorphisms Based on Physiochemical Features," International Journal of Bioscience, Biochemistry and Bioinformatics vol. 1, no. 2, pp. 102-108, 2011.