Dataset Information

Optimal settings of mass spectrometry open search strategy for higher confidence

ABSTRACT: More than half of the mass spectra could not be identified in most proteome mass spectrometry experiments. Various modifications have been considered as a major reason. Open search strategy was then introduced to solve this problem, however, lacking thorough quality assessment using independent information. Here, we used the “Suspicious Discovery Rate (SDR)” based on the translatome sequencing (RNC-seq) as an independent source to assess the proteome open search strategy. We found that the open search strategy increased the spectra utilization with the cost of increasing suspicious identifications that lacks translation evidence. We further suggested that restricting the peptide FDR below 0.1% would efficiently control the suspicious identifications of open search methods and enhanced the confidence of the peptide identification with modifications than the narrow window search. These results facilitated the proper use of open search methods for higher quality of proteome identifications with the information of post-translational modifications and single amino acid polymorphisms

ORGANISM(S): Homo sapiens

PROVIDER: GSE112623 | GEO | 2018/10/04

REPOSITORIES: GEO

ACCESS DATA

Similar Datasets

Project description:In recent years, high-throughput technologies have contributed to development a more precise picture of the human proteome. However, 2,129 proteins remain listed as missing proteins (MPs) in the newest neXtProt release (2019-02). The main reasons for MPs are a low abundance, low-molecular-weight (LMW), unexpected modifications, membrane characteristics, etc. Moreover, more than 50% of the MS/MS data have not been successfully identified in shotgun proteomics. Open-pFind, an efficient open search engine, recently released by the pFind group in China, presents an opportunity to identify these buried MPs in complex samples. Proteins and potential MPs were identified using Open-pFind and three other search engines to compare their performance and efficiency with three large-scale datasets digested by different enzymes. Our results demonstrated that Open-pFind identified 29.9-47.5% more peptide-spectrum matches (PSMs), 48.0-63.9% more peptides sequences (with modifications) and 22.7-38.1% more peptide sequences (regardless of modifications) than those identified by the second-best search engine. As a result, Open-pFind detected 7.5-19.3% more candidate MPs than those by the second-best search engine. In total, 5 (PE2) of the 150 candidate MPs identified by Open-pFind were verified from two unique peptides containing more than 9 amino acids (AA) by using spectrum theoretical prediction with pDeep, and synthesized peptide matching with pBuild, after spectrum quality analysis, isobaric post-translational modification, and single amino acid variant (SAAV) filtering. These five verified MPs can be ranked in the PE1 level. In addition, three other candidate MPs were verified with two unique peptides (one peptide containing more than 9 AA) and the other containing only 8 AA), which were slightly lower than the criteria listed by C-HPP, and required additional verification information. More importantly, unexpected modifications were detected in these MPs. Another 141 MPs were listed as candidates, but required additional verification information.

Dataset Information

Optimal settings of mass spectrometry open search strategy for higher confidence

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure