Guardat en:
| Autors principals: | , |
|---|---|
| Format: | Recurso digital |
| Idioma: | |
| Publicat: |
Zenodo
2025
|
| Accés en línia: | https://doi.org/10.5281/zenodo.16877509 |
| Etiquetes: |
Afegir etiqueta
Sense etiquetes, Sigues el primer a etiquetar aquest registre!
|
Taula de continguts:
- <p>This Evaluator requests an <strong>interaction matrix readout for chromatin conformation in H1 cells</strong> (homo sapiens). It can be used with Predictors such as <strong>ORCA</strong> that have been trained on Hi-C data.</p> <p>Included in <code>Test_setChr9_Evaluator.sif</code> :</p> <ul> <li>Scripts to process the data and connect to predictors in the GAME API</li> <li>ORCA evaluation support scripts from <a href="https://github.com/jzhoulab/orca.git">jzhoulab/orca</a></li> <li>Selene scripts from <a href="https://github.com/kathyxchen/selene.git">kathyxchen/selene</a></li> <li>Target values (large files) and hg38 reference genome</li> </ul> <p><strong>Prediction request details</strong></p> <ul> <li>This evaluator requests predictions for 1M bp long sequences spanning chromosome 9 (138 sequences total) using hg38 annotations</li> </ul> <p><code>/evaluator_data</code> folder contents:</p> <ul> <li><strong>JSON request file</strong> <ul> <li>Contains request metadata in the GAME API format</li> <li>Stores <strong>sequence start coordinates</strong> instead of full sequences to save space</li> <li>These coordinates are used by the main evaluator script to extract sequences</li> </ul> </li> <li>The folder also contains a python script to create the json file in case users want more details to reproduce the input data</li> </ul> <p>How to run:</p> <div><code>apptainer run --containall -B /path_to/evaluator_data/:/evaluator_data -B /path_to/prediction_folder/:/predictions Test_setChr9_Evaluator.sif HOST PORT /predictions</code></div> <p>Evaluation details:</p> <ul> <li>The 1Mbp long sequences will be pulled at runtime </li> <li>The Evaluator loads target Hi-C matrices and computes the Pearson r correlation between predictions and the measurements.</li> <li>This Evaluator can only evaluate the predictions from 1Mbp long sequences from the 1M resolution ORCA predictor for H1-ESC.</li> <li>Chromosome 9 is one of ORCA's test set chromosomes</li> <li>Matrices with all NA predictions are excluded from correlation calculations</li> </ul> <p>Additional information regarding ORCA and its evaluation can be found here: <br><a href="https://github.com/jzhoulab/orca">https://github.com/jzhoulab/orca</a></p> <p><a href="https://github.com/jzhoulab/orca_manuscript/blob/main/Model_performance.ipynb">https://github.com/jzhoulab/orca_manuscript/blob/main/Model_performance.ipynb</a></p> <p>Additional information regarding the API can be found here: <a href="https://github.com/de-Boer-Lab/Genomic-API-for-Model-Evaluation">https://github.com/de-Boer-Lab/Genomic-API-for-Model-Evaluation</a></p>