Monitoring HCPs: A Comparative Analysis of MS/MS Search Algorithms
Somar Khalil discusses label-free proteomics for Host Cell Protein monitoring, evaluating MS/MS tools and their performance metrics.
About This Webinar
Monitoring Host-Cell Proteins: A Comparative Analysis of MS/MS Search Algorithms
Presented by Somar Khalil, Principal Scientist, GSK Vaccines in November 2024
Monitoring host cell proteins (HCPs) is crucial for ensuring product safety and quality in biopharmaceutical development. In this webinar, Somar presents the application of label-free shotgun proteomics for HCP monitoring, benchmarking six prominent MS/MS search tools—Mascot, MaxQuant, SpectroMine, FragPipe, Byos, and PEAKS—utilizing trapped ion mobility spectrometry and ddaPASEF. The study employs Bayesian inference as a robust framework for evaluating tool performance, incorporating Markov Chain Monte Carlo modeling, Bayesian FDR control, and decision theory to balance differential abundance accuracy with uncertainty.
The webinar will highlight the strengths and trade-offs of each search tool, focusing on key metrics such as peptide and protein identifications, data extraction precision, fold change accuracy, linearity, and quantification trueness.
On utility scoring: "Byos followed by Spectromine and Mascot ranked highest, validating their robust performance across spike levels." 24:56
On linearity: "Byos demonstrated the highest linearity, closely followed by SpectroMine and Mascot, suggesting these tools performed well in capturing expected proportional changes." 27:06
On trueness: "Byos exhibited the highest trueness with the lowest mean absolute values, suggesting a superior accuracy in representing true values." 27:53
On overall performance: "Byos stands out as the most consistent and accurate tool, followed by SpectroMine and Mascot, which demonstrated comparable levels of linearity and trueness." 28:20
On protein quantification: "Byos performed robustly, increasing from three hundred fifty up to eight hundred forty proteins" 18:35