Abstract:Host cell proteins (HCPs) are critical quality attributes that can impact the safety, efficacy, and quality of biotherapeutics. Label-free shotgun proteomics is a vital approach for HCP monitoring, yet the choice of tandem mass spectrometry (MS/MS) search algorithms directly influences identification depth and quantification reliability. In this study, six prominent MS/MS search tools, Mascot, MaxQuant, SpectroMine, FragPipe, Byos, and PEAKS, were systematically benchmarked for their performance on complex samples spiked with isotopically labeled proteins from Chinese hamster ovary cells, using trapped ion mobility spectrometry and parallel accumulation-serial fragmentation in data-dependent acquisition mode. Key performance metrics, including peptide and protein identifications, data extraction precision, fold-change accuracy, linearity, and measurement trueness, were evaluated. A Bayesian modeling framework with Hamiltonian Monte Carlo sampling was employed to robustly estimate fold-change means and variances, alongside local false discovery rates through posterior probability calibration. Bayesian decision theory, implemented via expected utility maximization, was used to balance accuracy against posterior uncertainty, providing a probabilistic assessment of each tools performance. Through this cumulative analysis, variability across tools was observed: some excelled in identification sensitivity and protein coverage, others in quantitative accuracy with minimal bias, and a few offered balanced performance across metrics. This study establishes a rigorous, data-driven framework for tool benchmarking, delivering insights for selecting MS/MS tools suited to HCP monitoring in biopharmaceutical development.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the performance differences of different tandem mass spectrometry (MS/MS) search algorithms in monitoring host cell proteins (HCPs) in label - free quantitative proteomics. Specifically, the paper evaluates the performance of six commonly - used MS/MS search tools (Mascot, MaxQuant, SpectroMine, FragPipe, Byos, and PEAKS) in complex samples by comparing them, especially the performance of these tools in peptide and protein identification, data extraction accuracy, fold - change accuracy, linearity, and measurement authenticity. ### Background Host cell proteins (HCPs) are endogenous impurities in the production process of recombinant therapeutic drugs and may affect the safety, efficacy, and quality of biopharmaceuticals. Therefore, strict monitoring and accurate quantification of HCPs are crucial. Label - free quantification (LFQ) has become a powerful tool in shotgun proteomics for identifying and quantifying HCPs, but low - abundance HCPs and a wide dynamic range pose challenges for detection. ### Research Objectives 1. **Evaluate the performance of different MS/MS search tools**: Systematically compare six commonly - used MS/MS search tools to evaluate their identification and quantification capabilities in complex samples. 2. **Optimize HCPs monitoring methods**: Utilize trapped - ion mobility spectrometry (TIMS) and data - dependent acquisition - parallelizable sequential fragmentation (ddaPASEF) techniques to improve the detection sensitivity and coverage of HCPs. 3. **Provide data - driven tool selection recommendations**: Establish a rigorous framework through Bayesian inference and Hamiltonian Monte Carlo (HMC) sampling to evaluate the performance of each tool and provide a selection basis for HCP monitoring in biopharmaceutical development. ### Methods - **Sample preparation**: Use stable isotope - labeled Chinese hamster ovary (CHO) proteins to prepare antigen - drug substance (DS) samples at different concentrations. - **Mass spectrometry analysis**: Perform liquid chromatography - mass spectrometry analysis in TIMS and ddaPASEF modes. - **Data processing**: Use six different MS/MS search tools to process the raw data and evaluate key performance indicators such as peptide and protein identification and data extraction accuracy, fold - change accuracy, etc. - **Statistical analysis**: Use Bayesian inference and HMC sampling to evaluate the performance of each tool and balance accuracy and posterior uncertainty through Bayesian decision theory. ### Results - **Protein and peptide identification**: PEAKS performs excellently in protein identification, while Byos leads in peptide identification. - **Data extraction accuracy**: FragPipe performs best in data extraction accuracy, followed by SpectroMine. - **Fold - change accuracy**: In low - concentration samples, FragPipe, Byos, and SpectroMine perform excellently; in high - concentration samples, SpectroMine has the highest fold - change accuracy. - **Linearity and measurement authenticity**: Byos performs best in linearity and measurement authenticity at the protein level, followed by PEAKS. ### Conclusion Through comprehensive evaluation, the study found that different MS/MS search tools have significant differences in identification sensitivity, quantitative accuracy, and other performance indicators. These results provide a data - driven selection basis for the biopharmaceutical industry and help optimize HCPs monitoring methods.

Comparative analysis of MS/MS search algorithms in label-free shotgun proteomics for monitoring host-cell proteins using trapped ion mobility and ddaPASEF

Comparative evaluation of label-free quantification strategies.

Optimization of Search Engines and Postprocessing Approaches to Maximize Peptide and Protein Identification for High-Resolution Mass Data

Comprehensive host cell proteins profiling in biopharmaceuticals by a sensitivity enhanced mass spectrometry strategy using TMT-labeling and signal boosting

Systematic Assessment of Survey Scan and Ms2-Based Abundance Strategies for Label-Free Quantitative Proteomics Using High-Resolution Ms Data

A fully automated system with online sample loading, isotope dimethyl labeling and multidimensional separation for high-throughput quantitative proteome analysis.

Label-Free Protein Quantification Using Lc-Coupled Ion Trap or Ft Mass Spectrometry: Reproducibility, Linearity, and Application with Complex Proteomes

Comparison of Database Searching Programs for the Analysis of Single-Cell Proteomics Data.

Performance Investigation Of Proteomic Identification By Hcd/Cid Fragmentations In Combination With High/Low-Resolution Detectors On A Tribrid, High-Field Orbitrap Instrument

Comparative Evaluation of Proteome Discoverer and FragPipe for the TMT-Based Proteome Quantification

Improving Peptide Identification with Single-Stage Mass Spectrum Peaks.

StatsPro: Systematic Integration and Evaluation of Statistical Approaches for Detecting Differential Expression in Label-Free Quantitative Proteomics.

Targeted Feature Detection for Data-Dependent Shotgun Proteomics

Staged-probability strategy of processing shotgun proteomic data to discover more functionally important proteins

Optimization And Evaluation Of Orbitrap Exploris 480 Mass Spectrometry For Quantitative Proteomics

High-quality and Robust Protein Quantification in Large Clinical/pharmaceutical Cohorts with IonStar Proteomics Investigation

MSstatsPTM: Statistical Relative Quantification of Posttranslational Modifications in Bottom-Up Mass Spectrometry-Based Proteomics

Isobaric Tagging and Data Independent Acquisition as Complementary Strategies for Proteome Profiling on an Orbitrap Astral Mass Spectrometer

Binomial probability distribution model-based protein identification algorithm for tandem mass spectrometry utilizing peak intensity information.

A Mass-Tolerant Database Search Identifies a Large Proportion of Unassigned Spectra in Shotgun Proteomics As Modified Peptides

Protein-Level Integration Strategy of Multiengine MS Spectra Search Results for Higher Confidence and Sequence Coverage