Quantifying the ∼75-95% of Peptides in DIA-MS Data Sets that Were Not Previously Quantified

Area of Science:

• Proteomics
• Analytical Chemistry
• Biotechnology

Background:

• Data-independent acquisition liquid-mass spectrometry (DIA-LC-MS) is a powerful technique for peptide quantification.
• Current algorithms face challenges in accurately quantifying peptides, especially those without pre-assigned sequences.
• There is a need for improved algorithms to maximize the information extracted from DIA-LC-MS data.

Purpose of the Study:

• To introduce GoldenHaystack (GH), a novel algorithm for enhanced peptide quantification in DIA-LC-MS data.
• To demonstrate GH's ability to quantify peptides regardless of sequence assignment.
• To compare GH's performance against existing DIA-MS algorithms like DIA-NN.

Main Methods:

• GoldenHaystack (GH) employs multi-partite matching of MS2 and MS1 ions across all samples.
• It separates and regroups MS ions into unique analyte quantifiable signatures (UAQS) to reduce noise.
• GH quantifies UAQS, with optional peptide identification using FASTA files post-quantification.

Main Results:

• GH quantifies and identifies known peptides with improved FDR accuracy compared to DIA-NN.
• GH quantifies a significant portion (∼75-95%) of typically unassigned peptides.
• GH demonstrates substantial speed improvements, running 40-200x faster than DIA-NN and 1-10x faster than LC-MS acquisition.

Conclusions:

• GoldenHaystack (GH) offers a significant advancement in DIA-LC-MS peptide quantification.
• GH enhances proteomic analysis by accurately quantifying both assigned and unassigned peptides.
• The algorithm's speed and accuracy make it a valuable tool for large-scale proteomics studies.