Quandenser+Triqler

Quandenser+Triqler improves label-free quantification in shotgun proteomics by applying a quantification-first workflow with unsupervised MS1/MS2 clustering to summarize analytes and by integrating uncertainty with Triqler's Bayesian model for more accurate differential protein quantification.

Key Features:

• Quantification‑first approach: Prioritizes quantification over identification by summarizing analytes before identity assignment, reversing traditional search-first workflows.
• Unsupervised MS1/MS2 clustering: Applies unsupervised clustering at both MS1 and MS2 levels to reduce data volume and summarize analytes for downstream analysis.
• Data reduction and search-time reduction: Summarizing analytes without immediate identification reduces the number of spectra for database searching and accelerates analysis.
• Enhanced sensitivity: Prevents discarding information during identification, improving detection of differentially abundant proteins and outperforming methods such as MaxQuant+Perseus across evaluated datasets.
• Error regulation with Triqler: Integrates multiple sources of uncertainty into a single combined quantification error using a Bayesian model to provide more accurate and reliable quantification.
• MaxQuant compatibility: Triqler can process MaxQuant results to enable reanalysis of existing datasets.

Scientific Applications:

• Label‑free quantification experiments: Enhances analysis of label-free shotgun proteomics datasets by improving quantification before identification.
• Differential protein expression analysis: Increases sensitivity and reliability in detecting differentially abundant proteins.
• Complex biological and clinical datasets: Recovers analytes that may be overlooked by traditional pipelines, supporting engineered and clinical/biological sample analysis.
• Biomarker discovery and disease mechanism studies: Supports discovery of biomarkers and investigation of disease mechanisms through improved quantification and error modeling.

Methodology:

Quandenser performs unsupervised clustering at MS1 and MS2 to summarize analytes prior to identification and reduce search space; Triqler integrates multiple sources of uncertainty into a single combined quantification error using a Bayesian model and can process MaxQuant results.