Informatics strategies for large-scale novel cross-linking analysis.

Area of Science:

• Proteomics
• Biochemistry
• Computational Biology

Background:

• Detecting protein interactions is a major technological challenge.
• Chemical cross-linking aids in identifying protein-protein interactions but faces complexity and heterogeneity issues.
• Mass spectrometry-identifiable cross-linkers, protein interaction reporters (PIRs), enable on-cell experiments but generate complex data.

Purpose of the Study:

• To develop computational tools for interpreting complex data from PIR experiments.
• To introduce the X-links program for differentiating PIR product types.
• To estimate false discovery rates (FDR) for PIR product and peptide identification.

Main Methods:

• Development of the X-links software for PIR product type differentiation.
• Monte Carlo simulations to estimate FDR for PIR product identification using precursor and released peptide masses.
• Peptide identification calculations based on accurate masses and database complexity.

Main Results:

• The X-links program enables PIR product type differentiation.
• Simulations indicate a low FDR (approx. 12%) for PIR product type identification with 10 ppm accuracy and 100 peptides.
• Using a reduced database (Shewanella oneidensis MR-1, 367 proteins) significantly decreased the expected FDR for peptide identification.

Conclusions:

• The developed informatics capabilities and X-links software facilitate the interpretation of PIR experiment data.
• PIRs offer a viable method for large-scale protein-protein interaction identification with manageable FDR.
• Computational approaches, including database reduction, are crucial for accurate proteomic analysis.