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

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. The data were acquired 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 (FC) accuracy, linearity, and measurement trueness, were evaluated. A Bayesian modeling framework with Hamiltonian Monte Carlo sampling was employed to robustly estimate FC 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 and provide a probabilistic assessment of each tool’s performance. Through this cumulative analysis, variability across tools was observed: Byos and SpectroMine excelled in quantitative accuracy with minimal bias, FragPipe provided high precision and quantifiability, PEAKS offered deep protein coverage, Mascot showed strong trueness, and MaxQuant exhibited moderate identification performance with greater variability at lower spike levels. This study establishes a rigorous, data-driven framework for tool benchmarking and offers guidance for selecting MS/MS tools suited to HCP monitoring in biopharmaceutical development.