报告时间：2018年12月26日（星期三）上午10:00

报告地点：生物楼学术报告厅

报告人：陈镇霖博士，中国科学院计算技术研究所

　　报告摘要：

　　Cross-linking mass spectrometry (CXMS) is a powerful technique to discover protein-protein interactions, potentially at a proteome scale. However, currently available software tools for CXMS data analysis still suffer from poor speed and credibility, especially when searching a large protein database. In this paper, we propose a search engine, pLink 2, for proteome-scale identification of cross-linked peptides with higher speed and credibility. With a novel two-stage open search strategy facilitated by fragment indexing, pLink 2 is on average 40 times faster than pLink 1 and 3~10 times faster than Kojak. Furthermore, four new analysis methods, using simulated datasets, synthetic datasets, 15N metabolically labeled datasets, and entrapment databases, have been designed to evaluate the credibility of ten state-of-the-art search engines, and this systematic evaluation showed that pLink 2 achieved the highest precision and sensitivity, by a large margin at proteome scales. Finally, reanalysis of four published cell lysate datasets using pLink 2 took only a fraction of the time used by pLink 1, with up to 27% more cross-linked sites identified. The new search engine pLink 2 and the systematic evaluation methods are well positioned to become new benchmarks for cross-link identification research.

　　联系人：1810组 张丽华

　　联系电话：89720