Progress on the development of prediction tools for detecting disease causing mutations in proteins

Area of Science:

• Biochemistry and Molecular Biology
• Computational Biology
• Genetics

Background:

• Protein mutations alter structure, stability, binding, and function, potentially causing diseases.
• Understanding these mutation effects is crucial for disease mechanism insights and therapeutic strategies.
• Numerous methods exist to identify pathogenic mutations, driving research in this area.

Purpose of the Study:

• To review the development of computational prediction methods for identifying disease-causing protein mutations.
• To provide an overview of existing databases and sequence/structure-based features for mutation prediction.
• To discuss the application of machine learning, deep learning, and large language models in this field.

Main Methods:

• Database compilation of disease-causing mutations.
• Analysis of sequence- and structure-based features for predictive modeling.
• Discussion of computational tools, including machine learning, deep learning, and large language models.
• Focus on specific disease areas (cancer, neurodegenerative, infectious) and protein types (membrane proteins).

Main Results:

• Overview of current databases and prediction features.
• Highlighting advances in machine learning, deep learning, and large language models for mutation effect prediction.
• Emphasis on predicting mutations in key disease targets and membrane proteins.
• Identification of computational resources for mutation effect analysis.

Conclusions:

• Computational methods are advancing for predicting disease-causing mutations.
• Further improvements are needed to enhance the accuracy and scope of existing prediction tools.
• This review provides a resource for understanding and utilizing computational approaches in mutation research.