Retention Time Prediction Using Neural Networks Increases Identifications in Crosslinking Mass Spectrometry

Abstract

Crosslinking mass spectrometry (Crosslinking MS) has developed into a robust technique that is increasingly used to investigate the interactomes of organelles and cells. However, the incomplete and noisy information in the spectra limits the numbers of protein-protein interactions (PPIs) that can be confidently identified. Here, we successfully leveraged chromatographic retention time (RT) information to aid the identification of crosslinked peptides from spectra. Our Siamese machine learning model xiRT achieved highly accurate RT predictions of crosslinked peptides in a multi-dimensional separation of crosslinked E. coli lysate. We combined strong cation exchange (SCX), hydrophilic strong anion exchange (hSAX) and reversed-phase (RP) chromatography and reached R^2 0.94 in RP and a margin of error of 1 fraction for hSAX in 94%, and SCX in 85% of the predictions. Importantly, supplementing the search engine score with retention time features led to a 1.4-fold increase in PPIs at a 1% false discovery rate. We also demonstrate the value of this approach for the more routine analysis of a crosslinked multiprotein complexes. An increase of 1.7-fold in heteromeric crosslinked residue-pairs was achieved at 1% residue-pair FDR for Fanconi anaemia monoubiquitin ligase complex, solely using reversed-phase RT. Retention times are a powerful complement to mass spectrometric information to increase the sensitivity of Crosslinking MS analyses.

Improved peptide backbone fragmentation is the primary advantage of MS-cleavable crosslinkers

“…S6), indicating that self-CSMs are approaching exhaustive coverage at the given experimental detection limit. Similar results were seen when including retention time data of heteromeric and self-CSMs 20 .…”

“…This bases on the assumption that not knowing the individual peptide masses before the search results in the need for exhaustive combination of all peptides in the database and thus an explosion of the search space. However, there are multiple large-scale studies that have successfully employed a non-cleavable crosslinker despite these assumptions 4,12,20,21 . These are based on a detailed understanding of how crosslinked peptides fragment 22 , that offered a computational solution to knowing the individual peptide masses which was then implemented in the search algorithm xiSEARCH 3 .…”

Accurate Retention Time Prediction Based on Monolinked Peptide Information to Confidently Identify Cross-Linked Peptides

“…35−37 We just noticed that a recent paper also used retention time prediction to increase identification of cross-linked peptides. 38 Herein, we found that the monolinked peptides carried out the information on the retention time of the corresponding cross-linked pairs, and the retention time of cross-linked peptides can be well predicted by the use of the retention time of the monolinked peptides. Unlike linear peptide prediction, the training data set and the test data set in this study can be generated from same LC-MS/ MS data set, which is more like internal calibration and will ensure that the prediction of cross-linked peptides is always accurate.…”

Label-free visual proteomics: Coupling MS- and EM-based approaches in structural biology

“…By pulling the crosslinked peptide pairs from the background, significantly higher depth of analysis can be achieved. Further optimizations are currently being explored with, for example, extensive fractionation (Lenz et al, 2021), advanced MS setups (Steigenberger et al, 2020), tailored data acquisition approaches (Hauri et al, 2019;Giese et al, 2021), and more robust reporting in terms of false-positive identifications (Lenz et al, 2021). A combination of these modified reagents, technological advancements, and optimization of reagents' membrane permeability will undoubtedly further improve proteome-wide in situ crosslinking results.…”