A Review on the Computational Tools for Predicting the Functional Impact of Missense Mutation

Abstract

Advances in genomic sequencing have left us with millions of genetic variants, the highest percentage of which constitutes missense mutation. Missense mutations are responsible for more than 50% of human-inherited diseases. The accurate characterization of which mutations lead to disease is an indispensable asset for the genomic era. Besides the endless potential for precision and personalized medicine, is the opportunity for timely and appropriate clinical interventions. To this end, the development of tools for predicting the functional impact of missense mutation remains an active area of research. In this review, we present a brief introduction and discussion on the state-of-the-art computational tools for predicting the functional impact of missense mutation. The focus is on the principles, features and methods employed by each tool. The methods employed were grouped into three; -Sequence-based methods, Machine learning methods and Graph-based methods. It was found that graph-based methods were able to capture the spatial structure of the protein in addition to features used by sequence-based and machine learning methods. Besides, graph-based models create opportunities for structure comparison of the wildtype and mutant protein by leveraging on the emerging field of graph representation learning.