Benchmarking Spectral Library and Database Search Approaches for Metaproteomics Using a Ground-Truth Microbiome...

Area of Science:

• Microbiology
• Proteomics
• Bioinformatics

Background:

• Metaproteomics is crucial for understanding microbiome functional interactions.
• Challenges exist in analyzing metaproteomics data, particularly in database searching.
• Evaluating different search strategies is essential for accurate microbiome analysis.

Purpose of the Study:

• To compare spectral library searching (Scribe) against database searching (MaxQuant, FragPipe) for metaproteomics.
• To assess the impact of search methods on protein detection, peptide-spectral match quality, and quantification.
• To provide guidance for optimizing metaproteomics data analysis.

Main Methods:

• Utilized a ground-truth dataset for metaproteomics analysis.
• Analyzed data-dependent acquisition mass spectrometry (DDA-MS) data.
• Employed MaxQuant, FragPipe, and Scribe with Prosit predicted spectral libraries against curated FASTA databases.

Main Results:

• Scribe detected more proteins at a 1% false discovery rate (FDR) than MaxQuant or FragPipe.
• FragPipe identified more peptides verified by PepQuery.
• Scribe demonstrated superior detection of low-abundance proteins and more accurate quantification of microbial community composition.

Conclusions:

• Spectral library searching offers advantages for metaproteomics analysis of DDA-MS data.
• Scribe provides improved sensitivity and accuracy for microbiome composition analysis.
• This study offers valuable insights for researchers optimizing metaproteomics workflows.