Comparative analysis of web-based programs for single amino acid substitutions in proteins

Area of Science:

• Biophysics
• Genomics
• Computational Biology

Background:

• Single amino acid substitutions in proteins can alter structure and function, leading to complex diseases.
• Analyzing these protein mutations is vital for understanding disease mechanisms.
• Numerous bioinformatics tools exist to predict the impact of mutations on human proteins.

Purpose of the Study:

• To systematically compare the prediction efficiency of commonly used protein mutational analysis programs.
• To evaluate 10 sequence-based and 5 structure-based tools for their accuracy in predicting pathogenic single point mutations.

Main Methods:

• Performed extensive mutational analyses using 15 different prediction tools.
• Utilized previously identified pathogenic single point mutations from five distinct proteins for evaluation.
• Compared prediction outcomes (scores indicating deleterious probability) from each tool.

Main Results:

• Identified PolyPhen2, PROVEAN, and PMut (sequence-based) as having high prediction accuracy.
• Found mCSM (structure-based) to demonstrate superior prediction accuracy.
• Sequence-based tools and structure-based tools showed varying degrees of prediction efficiency.

Conclusions:

• Sequence-based tools like PolyPhen2, PROVEAN, and PMut, along with the structure-based tool mCSM, exhibit strong prediction accuracy for pathogenic mutations.
• Employing multiple prediction programs utilizing different analytical approaches can significantly enhance the overall predictive power for protein mutation effects.