Spatially interacting phosphorylation sites and mutations in cancer

Area of Science:

• Proteomics
• Bioinformatics
• Structural Biology

Background:

• Mass spectrometry generates large proteomics datasets with numerous phosphorylation sites requiring prioritization.
• Identifying functional relationships between phosphorylation and cancer mutations is crucial for understanding disease mechanisms.

Purpose of the Study:

• To develop a bioinformatics tool (HotPho) for prioritizing phosphorylation sites.
• To systematically discover 3D co-clustering of phosphosites and cancer mutations on protein structures.
• To identify novel functional mutations and cancer-associated phosphosites.

Main Methods:

• Development of the HotPho bioinformatics tool.
• Systematic 3D co-clustering analysis of phosphosites and cancer mutations on protein structures.
• Validation of co-clustering sites in patient samples and existing literature.

Main Results:

• HotPho identified 474 hybrid clusters with 1255 co-clustering phosphosites.
• Novel co-clustering sites adjacent to mutations were found, missed by linear approaches (e.g., RET p.S904/Y928, HRAS/KRAS p.Y96).
• Approximately 300 co-clustering phosphosites are validated in patient samples or implicated in cancer.

Conclusions:

• Systematic 3D clustering analysis highlights nearly 3,000 likely functional mutations.
• Over 1000 cancer phosphosites are identified for further investigation.
• This approach offers potential clinical relevance for cancer research.