mCSM: predicting the effects of mutations in proteins using graph-based signatures

Area of Science:

• Genomics and Bioinformatics
• Molecular Biology
• Computational Biology

Background:

• Mutations introduce genomic diversity, driving evolution.
• Missense mutations in structural genes can alter protein stability and interactions, impacting organismal fitness.
• Predicting mutation impact is vital for understanding genetic diseases.

Purpose of the Study:

• To develop a novel computational approach, mCSM, for predicting the effects of missense mutations.
• To evaluate mCSM's performance in predicting changes in protein stability and interactions.
• To assess the utility of mCSM in disease-related mutation analysis.

Main Methods:

• Utilized graph-based signatures encoding atomic distance patterns to represent protein residue environments.
• Trained predictive models using these graph-based signatures.
• Applied mCSM to predict stability changes in mutations within the p53 tumor suppressor protein.

Main Results:

• mCSM demonstrates comparable or superior performance to existing methods.
• Mutation impact is correlated with atomic-distance patterns surrounding amino acid residues.
• mCSM successfully predicted stability changes for various mutations in the p53 protein.

Conclusions:

• mCSM is a valuable tool for predicting missense mutation impacts on protein stability and interactions.
• The method shows promise for applications in understanding disease progression.
• The approach highlights the utility of graph-based signatures in computational biology.