A bioinformatics workflow for variant peptide detection in shotgun proteomics.

Area of Science:

• Proteomics
• Genomics
• Bioinformatics

Background:

• Shotgun proteomics typically uses protein databases lacking variant information, hindering variant peptide identification.
• Existing protein sequence databases do not encompass known protein variations, limiting comprehensive proteomic analysis.

Purpose of the Study:

• To develop a data analysis workflow for identifying variant peptides in shotgun proteomics data.
• To create an enhanced protein sequence database incorporating known coding variations.

Main Methods:

• Constructed a protein sequence database integrating cancer-related variants (Cancer Proteome Variation Database) and non-cancer variants (dbSNP).
• Developed a shotgun proteomics data analysis workflow for variant peptide detection with a modified false discovery rate estimation.
• Validated the workflow using colorectal cancer cell lines (SW480, RKO, HCT-116) and tumor specimens.

Main Results:

• Identified 81 variant peptides in colorectal cancer cell lines, with 23 of 26 variants confirmed by genomic sequencing.
• Detected 204 distinct variant peptides in tumor specimens, including five with known cancer-related mutations.
• Demonstrated that each tumor specimen exhibited a unique mutation pattern, suggesting potential for personalized medicine.

Conclusions:

• The developed workflow effectively leverages genomic data for variant peptide detection in proteomics.
• The approach shows promise for identifying cancer-specific mutations and advancing personalized cancer medicine.
• The workflow is compatible with major database search engines (Sequest, Mascot, X!Tandem, MyriMatch).