References

Cite xiRT

Giese, S.H., Sinn, L.R., Wegner, F. et al. Retention time prediction using neural networks increases identifications in crosslinking mass spectrometry. Nature Communications 12, 3237 (2021). DOI: 10.1038/s41467-021-23441-0.

The publication describes the original scientific work; its existence does not validate every optional 2.0 model route or optimizer. Cite the exact software version, benchmark revision, and upstream embedding model when applicable.

Primary upstream documentation

• PyTorch installation and reproducibility notes.
• Setuptools pyproject configuration: static project metadata, modern build backend, license expressions.
• Material for MkDocs and Python mkdocstrings.
• Ruff configuration.
• MLflow tracking, MlflowClient, and local server CLI.
• PyPI trusted publishing and security model.
• GitHub build provenance action.
• Official GitHub Pages custom workflows.
• Predictive R² definition: this is not squared Pearson correlation.
• ESM upstream implementation and model documentation. xiRT currently supports SDK 3.2.x; the upstream main branch may describe newer, incompatible APIs and different model terms. Consult the precise downloaded model card and versioned SDK sources before using or redistributing weights.

Review upstream documentation and license terms for the exact versions used in your environment. Package compatibility bounds are not a claim that arbitrary future SDK/model revisions are interchangeable.