Bioinformatics analysis of correlation between protein function and intrinsic disorder

Area of Science:

• Proteomics
• Bioinformatics
• Structural Biology

Background:

• Understanding the link between protein intrinsic disorder and molecular function is crucial for deciphering protein behavior.
• Previous studies analyzed this correlation using Swiss-Prot keywords, prompting further investigation with more comprehensive functional annotations.

Purpose of the Study:

• To investigate the correlation between protein intrinsic disorder and molecular function using the CAFA3 dataset with Gene Ontology (GO) terms.
• To compare findings with a previous study that used Swiss-Prot keywords.
• To develop a novel method for mapping Swiss-Prot keywords to GO terms.

Main Methods:

• Utilized the CAFA3 challenge training dataset, annotating over 66,000 proteins with GO terms.
• Employed the PONDR VSL2b predictor to classify proteins as intrinsically disordered or ordered.
• Associated GO terms with Swiss-Prot function keywords and analyzed correlations.

Main Results:

• Identified significant numbers of both order-related and disorder-related Swiss-Prot keywords and GO terms.
• Found 746 GO terms to be highly order-related and 564 highly disorder-related.
• Observed consistent results between GO terms and Swiss-Prot keywords when mappings were possible, with a prevalence of disorder-related binding functions in the current dataset.

Conclusions:

• Protein intrinsic disorder is strongly correlated with specific molecular functions, particularly binding functions.
• The study highlights the utility of GO terms for a more detailed analysis of protein function-disorder relationships.
• A new method for mapping Swiss-Prot keywords to GO terms was proposed, enhancing functional annotation comparisons.