STAVER: A Standardized Benchmark Dataset-Based Algorithm for Effective Variation Reduction in Large-Scale DIA-MS Data

This project focuses on developing and applying STAVER, an innovative DIA algorithm designed to eliminate non-biological noise and variability from the large-scale DIA-MS study dataset analyses. STAVER is a flexible framework that utilizes prior knowledge regarding peptide separation coordinates (RT) and fragment ion intensities from the standard benchmark datasets, which effectively mitigates non-biological noise potential during library searches, enhancing spectrum identifications and protein quantification accuracy. Furthermore, the robustness and broad applicability of STAVER were validated in multiple large-scale DIA datasets from different platforms and laboratories, demonstrating significantly improved precision and reproducibility of protein quantification. It facilitates the comparative and integrative analysis of DIA datasets across different platforms and laboratories, enhancing the consistency and reliability of findings in clinical research. The project aims to promote the adoption of hybrid library search and improve the sensitivity and quality of DIA proteomics data through the open-source STAVER software package.

本项目聚焦于开发与应用STAVER——一款创新性的数据非依赖采集（Data Independent Acquisition, DIA）算法，旨在从大规模数据非依赖采集质谱（Data Independent Acquisition Mass Spectrometry, DIA-MS）研究的数据集分析中去除非生物噪声与变异。STAVER是一款灵活的分析框架，其利用源自标准基准数据集的肽段分离坐标（保留时间，RT）与碎片离子强度先验知识，可在数据库检索阶段有效抑制潜在非生物噪声，进而提升谱图鉴定效能与蛋白质定量准确性。此外，STAVER的稳健性与广泛适用性已在多平台、多实验室的多组大规模DIA数据集上得到验证，结果显示其可显著提升蛋白质定量的精密度与重现性。该工具可支持不同平台、不同实验室间DIA数据集的比较与整合分析，提升临床研究中研究结果的一致性与可靠性。本项目旨在通过开源STAVER软件包，推广混合数据库检索方法的应用，并提升DIA蛋白质组学数据的分析灵敏度与数据质量。