Deep learning the collisional cross sections of the peptide universe from a million experimental values

Meier, Florian; Köhler, Niklas; Brunner, Andreas-David; Wanka, Jean-Marc H.; Voytik, Eugenia; Strauss, Maximilian T.; Theis, Fabian J.; Mann, Matthias

doi:10.1038/s41467-021-21352-8

Cited by 112 publications

(230 citation statements)

References 82 publications

Supporting

Mentioning

190

Contrasting

Order By: Relevance

“…Because TIMS is a gas phase separation technique primarily controlled by electric potentials, a simple linear alignment can be sufficient to account for changes in the gas flow causing drifts in the measured ion mobility value over time. Illustrating this point, we observed a median coefficient of variation of 0.4% in a dataset of 168 LC–TIMS–MS experiments acquired on multiple instruments and over several months ( 92 ). In that study, we also estimated that determining the mobility position of a precursor ion within ∼1% accuracy reduces the number of peptide candidates in a ±1.5 ppm mass window by a factor of 2 to 3.…”

Section: The “Perfect Data Cuboid” Generated By Timsmentioning

confidence: 75%

Trapped Ion Mobility Spectrometry and Parallel Accumulation–Serial Fragmentation in Proteomics

Meier

Park

Mann

2021

Molecular & Cellular Proteomics

Self Cite

122

View full text Add to dashboard Cite

Recent advances in efficiency and ease of implementation have rekindled interest in ion mobility spectrometry, a technique that separates gas phase ions by their size and shape and that can be hybridized with conventional LC and MS. Here, we review the recent development of trapped ion mobility spectrometry (TIMS) coupled to TOF mass analysis. In particular, the parallel accumulation–serial fragmentation (PASEF) operation mode offers unique advantages in terms of sequencing speed and sensitivity. Its defining feature is that it synchronizes the release of ions from the TIMS device with the downstream selection of precursors for fragmentation in a TIMS quadrupole TOF configuration. As ions are compressed into narrow ion mobility peaks, the number of peptide fragment ion spectra obtained in data-dependent or targeted analyses can be increased by an order of magnitude without compromising sensitivity. Taking advantage of the correlation between ion mobility and mass, the PASEF principle also multiplies the efficiency of data-independent acquisition. This makes the technology well suited for rapid proteome profiling, an increasingly important attribute in clinical proteomics, as well as for ultrasensitive measurements down to single cells. The speed and accuracy of TIMS and PASEF also enable precise measurements of collisional cross section values at the scale of more than a million data points and the development of neural networks capable of predicting them based only on peptide sequences. Peptide collisional cross section values can differ for isobaric sequences or positional isomers of post-translational modifications. This additional information may be leveraged in real time to direct data acquisition or in postprocessing to increase confidence in peptide identifications. These developments make TIMS quadrupole TOF PASEF a powerful and expandable platform for proteomics and beyond.

show abstract

Section: The “Perfect Data Cuboid” Generated By Timsmentioning

confidence: 75%

Trapped Ion Mobility Spectrometry and Parallel Accumulation–Serial Fragmentation in Proteomics

Meier

Park

Mann

2021

Molecular & Cellular Proteomics

Self Cite

122

View full text Add to dashboard Cite

show abstract

“…As the machine learning field continues to evolve and more tools become available, more timsTOF-specific properties could be predicted and incorporated as additional information next to the spectral intensity and retention time described in this study. For example, a newly published model (Meier et al, 2021), based on a deep recurrent neural network trained with timsTOF data, can now predict the collisional cross section (CCS) values for any peptide. Since these values can be derived from the ion mobility, this feature is a promising characteristic to include.…”

Section: Discussionmentioning

confidence: 99%

Ion Mobility Coupled to a Time-of-Flight Mass Analyzer Combined With Fragment Intensity Predictions Improves Identification of Classical Bioactive Peptides and Small Open Reading Frame-Encoded Peptides

Peeters

Baggerman

Gabriels

et al. 2021

Front. Cell Dev. Biol.

View full text Add to dashboard Cite

Bioactive peptides exhibit key roles in a wide variety of complex processes, such as regulation of body weight, learning, aging, and innate immune response. Next to the classical bioactive peptides, emerging from larger precursor proteins by specific proteolytic processing, a new class of peptides originating from small open reading frames (sORFs) have been recognized as important biological regulators. But their intrinsic properties, specific expression pattern and location on presumed non-coding regions have hindered the full characterization of the repertoire of bioactive peptides, despite their predominant role in various pathways. Although the development of peptidomics has offered the opportunity to study these peptides in vivo, it remains challenging to identify the full peptidome as the lack of cleavage enzyme specification and large search space complicates conventional database search approaches. In this study, we introduce a proteogenomics methodology using a new type of mass spectrometry instrument and the implementation of machine learning tools toward improved identification of potential bioactive peptides in the mouse brain. The application of trapped ion mobility spectrometry (tims) coupled to a time-of-flight mass analyzer (TOF) offers improved sensitivity, an enhanced peptide coverage, reduction in chemical noise and the reduced occurrence of chimeric spectra. Subsequent machine learning tools MS2PIP, predicting fragment ion intensities and DeepLC, predicting retention times, improve the database searching based on a large and comprehensive custom database containing both sORFs and alternative ORFs. Finally, the identification of peptides is further enhanced by applying the post-processing semi-supervised learning tool Percolator. Applying this workflow, the first peptidomics workflow combined with spectral intensity and retention time predictions, we identified a total of 167 predicted sORF-encoded peptides, of which 48 originating from presumed non-coding locations, next to 401 peptides from known neuropeptide precursors, linked to 66 annotated bioactive neuropeptides from within 22 different families. Additional PEAKS analysis expanded the pool of SEPs on presumed non-coding locations to 84, while an additional 204 peptides completed the list of peptides from neuropeptide precursors. Altogether, this study provides insights into a new robust pipeline that fuses technological advancements from different fields ensuring an improved coverage of the neuropeptidome in the mouse brain.

show abstract

“…In all these applications the challenge of overfitting mentioned earlier needs to be kept in mind. There are many more DL-based applications in proteomics, such as collisional cross section (CCS) prediction for ion mobility mass spectrometers (Meier et al, 2021) and a wide range these have recently been reviewed (Wen et al, 2020).…”

Section: Ll Ai In the Proteomics Workflowmentioning

confidence: 99%

Artificial intelligence for proteomics and biomarker discovery

et al. 2021

Self Cite

View full text Add to dashboard Cite

There is an avalanche of biomedical data generation and a parallel expansion in computational capabilities to analyze and make sense of these data. Starting with genome sequencing and widely employed deep sequencing technologies, these trends have now taken hold in all omics disciplines and increasingly call for multi-omics integration as well as data interpretation by artificial intelligence technologies. Here, we focus on mass spectrometry (MS)-based proteomics and describe how machine learning and, in particular, deep learning now predicts experimental peptide measurements from amino acid sequences alone. This will dramatically improve the quality and reliability of analytical workflows because experimental results should agree with predictions in a multi-dimensional data landscape. Machine learning has also become central to biomarker discovery from proteomics data, which now starts to outperform existing best-in-class assays. Finally, we discuss model transparency and explainability and data privacy that are required to deploy MS-based biomarkers in clinical settings.

show abstract

Deep learning the collisional cross sections of the peptide universe from a million experimental values

Cited by 112 publications

References 82 publications

Trapped Ion Mobility Spectrometry and Parallel Accumulation–Serial Fragmentation in Proteomics

Trapped Ion Mobility Spectrometry and Parallel Accumulation–Serial Fragmentation in Proteomics

Ion Mobility Coupled to a Time-of-Flight Mass Analyzer Combined With Fragment Intensity Predictions Improves Identification of Classical Bioactive Peptides and Small Open Reading Frame-Encoded Peptides

Artificial intelligence for proteomics and biomarker discovery

Contact Info

Product

Resources

About