🧬 2024borcsok_ERCC2

This repository contains computational predictions supporting the paper:
"Quantitative functional profiling of ERCC2 mutations deciphers cisplatin sensitivity in bladder cancer"
by Judit Börcsök, Djavarshini Gopaul, Daphne Devesa-Serrano, Clémence Mooser, Nicolas Jonsson, Matteo Cagiada, Dag R. Stormoen, Maya N. Ataya, Brendan J. Guercio, Hristos Z. Kaimakliotis, Gopa Iyer, Kresten Lindorff-Larsen, Lars Dyrskjøt, Kent W. Mouw, Zoltan Szallasi, and Claus S. Sørensen.
📄 [Link to paper – to be added]

📚 Contents

This repository includes predictions and analysis files generated using:

• 🧪 Cagiada Functional Site Model
• 🧬 FunC-ESMs annotations with ClinVar mappings (2023–2024)

📁 File Structure

• ercc2_outputs/
  • test_residue_predictions.csv
  • test_variant_features.csv
  • test_variant_predictions.csv
  • figures/
  • test_GEMME_histogram.png
  • test_RaSP_histogram.png
  • test_residue_class.png
  • test_variant_map.png
  • test_WCN_histogram.png
• func-esms_P18074-ERCC2_clinvar2023_clinvar2024_mapped.txt

⚙️ Methods

🧪 Cagiada Functional Site Model

All predictions in ercc2_outputs were generated using the model from:

📎 Google Colab Notebook – Cagiada Functional Site Model

This model provides residue- and variant-level features including GEMME, RaSP, WCN, and classification outputs.

🧬 FunC-ESMs Data

The file func-esms_P18074-ERCC2_clinvar2023_clinvar2024_mapped.txt was based on data from:
"Decoding molecular mechanisms for loss of function variants in the human proteome"
by M. Cagiada, N. Jonsson, and K. Lindorff-Larsen (2024, bioRxiv).
📄 View paper on bioRxiv
💾 Download dataset via KU ERDA
➡️ Navigate to erda/P1/80/74/ to access the specific file used.

Clinical annotations from ClinVar (2023 and 2024) were mapped onto ERCC2 variants in this file for this project.

📌 Citation

If you use this repository or its data, please cite the associated publication above as well as the tools listed below:

🧪 The Cagiada Functional Site Model

• Cite this article:
  Cagiada, M., Bottaro, S., Lindemose, S. et al. Discovering functionally important sites in proteins. Nat Commun 14, 4175 (2023).
  🌐 https://doi.org/10.1038/s41467-023-39909-0
  
• 🧭 This model was used to generate residue- and variant‑level predictions via the Colab notebook.

🧬 FunC‑ESMs Dataset & Tools

• Cite this article:
  Cagiada, M., Jonsson, N., & Lindorff‑Larsen, K. Decoding molecular mechanisms for loss‑of‑function variants in the human proteome. bioRxiv (2025).
  🌐 https://doi.org/10.1101/2024.05.21.595203
  
• 💾 Data was obtained from the KU ERDA repository. Navigate to erda/P1/80/74/ to download the FunC‑ESMs file used in this project.
  📎 ERDA Repository (share link)

Clinical variant annotations from ClinVar (2023 & 2024) were mapped into the final .txt file (func‑esms_P18074‑ERCC2_clinvar2023_clinvar2024_mapped.txt).