Template proteogenomics: sequencing whole proteins using an imperfect database.

Area of Science:

• Proteomics
• Bioinformatics
• Genomics

Background:

• Standard database search algorithms for tandem mass spectra identification are limited to known peptide sequences.
• Existing methods struggle with mutated, alternatively spliced, or missing sequences in databases.
• De novo peptide identification lacks accuracy due to the absence of a database.

Purpose of the Study:

• To develop a novel proteogenomic approach, GenoMS, integrating database and de novo methods.
• To overcome limitations of existing algorithms in identifying peptides from variable protein sequences.
• To accurately reconstruct full protein sequences, particularly for complex molecules like antibodies.

Main Methods:

• GenoMS utilizes InsPecT to identify protein sequence templates (homologous or genomic).
• It recruits, aligns, and de novo sequences divergent or missing regions of the target protein.
• The approach was tested on reconstructing antibody sequences using multiple digests and proteases.

Main Results:

• GenoMS accurately reconstructed full antibody sequences (>98% accuracy) using a diverged protein database.
• Using the genome as a template achieved >97% accuracy for antibody sequence reconstruction.
• Demonstrated effectiveness in identifying peptides from mutated and alternatively spliced sequences.

Conclusions:

• GenoMS offers a robust proteogenomic solution for comprehensive protein sequence identification.
• The method successfully addresses challenges posed by protein sequence variations and database limitations.
• GenoMS significantly advances the accuracy and scope of proteomic data analysis, especially for complex protein families like immunoglobulins.