Format

Send to

Choose Destination
Proteomics. 2016 Aug;16(15-16):2238-45. doi: 10.1002/pmic.201500499. Epub 2016 May 27.

SAINTq: Scoring protein-protein interactions in affinity purification - mass spectrometry experiments with fragment or peptide intensity data.

Author information

1
Saw Swee Hock School of Public Health, National University of Singapore and National University Health System, Singapore, Singapore.
2
Department of Pathology, Yale University, New Haven, CT, USA.
3
Lunenfeld-Tanenbaum Research Institute, Sinai Health Service, Ontario, Canada.
4
Department of Molecular Genetics, University of Toronto, Ontario, Canada.
5
Institute of Molecular and Cell Biology, Agency for Science, Technology and Research, Singapore, Singapore.

Abstract

SAINT (Significance Analysis of INTeractome) is a probabilistic method for scoring bait-prey interactions against negative controls in affinity purification - mass spectrometry (AP-MS) experiments. Our published SAINT algorithms use spectral counts or protein intensities as the input for calculating the probability of true interaction, which enables objective selection of high-confidence interactions with false discovery control. With the advent of new protein quantification methods such as Data Independent Acquisition (DIA), we redeveloped the scoring method to utilize the reproducibility information embedded in the peptide or fragment intensity data as a key scoring criterion, bypassing protein intensity summarization required in the previous SAINT workflow. The new software package, SAINTq, addresses key issues in the interaction scoring based on intensity data, including treatment of missing values and selection of peptides and fragments for scoring each prey protein. We applied SAINTq to two independent DIA AP-MS data sets profiling the interactome of MEPCE and EIF4A2 and that of 14-3-3β, and benchmarked the performance in terms of recovering previously reported literature interactions in the iRefIndex database. In both data sets, the SAINTq analysis using the fragment-level intensity data led to the most sensitive detection of literature interactions at the same level of specificity. This analysis outperforms the analysis using protein intensity data summed from fragment intensity data that is equivalent to the model in SAINTexpress.

KEYWORDS:

Bioinformatics; Interaction; Protein-protein

PMID:
27119218
DOI:
10.1002/pmic.201500499
[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Wiley
Loading ...
Support Center