deeplc 4.0.0

DeepLC: Retention time prediction for (modified) peptides using Deep Learning.

---

DeepLC: Retention time prediction for peptides carrying any modification.

---

About DeepLC

DeepLC predicts retention times for peptides carrying any modification. It does this by leveraging a deep learning model based on atomic composition features. Starting with v4, DeepLC comes with a multitask pretrained model covering multiple LC setups, enabling accurate predictions out of the box. For best results on a specific dataset, predictions can be calibrated or fine-tuned using a small reference set of identified PSMs.

Citation

If you use DeepLC, please cite:

DeepLC can predict retention times for peptides that carry as-yet unseen modifications
Robbin Bouwmeester, Ralf Gabriels, Niels Hulstaert, Lennart Martens & Sven Degroeve
Nature Methods 18, 1363–1369 (2021) doi:10.1038/s41592-021-01301-5

If you use the transfer learning functionality, please also cite:

Retention time prediction improves proteomics database search and identification rates
Nature Communications (2026) doi:10.1038/s41467-026-68981-5

To replicate the results from this paper, use DeepLC v3.1.13. For regular use, we recommend the latest stable version.

Usage

Web application

A hosted web application is available at iomics.ugent.be/deeplc — no installation required.

Local graphical interface

Windows: download the one-click installer from the releases page.

Other platforms: install with GUI dependencies and launch as a desktop app or local web server:

pip install deeplc[gui]
deeplc gui            # opens in browser
deeplc gui --native   # opens as desktop window

Command line and Python API

pip install deeplc
deeplc predict peptides.tsv

from psm_utils.io import read_file
from deeplc import predict_and_calibrate

psm_list = read_file("peptides.tsv")
calibrated_rt = predict_and_calibrate(psm_list)

See the documentation for the full CLI reference, Python API, and input file format.

Related projects

• im2deep — Ion mobility / collisional cross section prediction using the same atomic composition approach
• MS²Rescore — Peptide identification rescoring that uses DeepLC retention time predictions as a rescoring feature
• iDeepLC — DeepLC variant using molecular descriptors to incorporate full molecular structure into predictions; better performance for some amino acid modifications.

Documentation

Full documentation at deeplc.readthedocs.io:

• Usage — CLI, Python API, input formats
• Prediction models — Model descriptions and training data
• Migrating from v3 — API and format changes
• Changelog
• Contributing