Result: nf-core/drugresponseeval: Drugresponseeval 1.1.0 - Humongous Zapdos

What's Changed Added #43 Preprint is out now! Linking it in the documentation. #42 Added authors and licenses to the python scripts. #43 Added --no_hyperparameter_tuning flag for quick runs without hyperparameter tuning: hpam_split takes this as argument #43 Added --final_model_on_full data flag: if True, a final/production model is saved in the results directory. If hyperparameter_tuning is true, the final model is tuned, too. The model can later be loaded using the implemented load functions of the drevalpy models. New process FINAL_SPLIT: splits the full dataset for each model class into train, validation, and optionally early stopping. This is done per model class and not overall because here, we no longer need across-model compatibility but want to train on the maximum amount of data (which might vary between models due to different feature availability) New process TUNE_FINAL_MODEL: trains the final model(s) with all hyperparameter combinations Added process EVALUATE_AND_FIND_MAX_FINAL: re-uses the EVALUATE_AND_FIND_MAX process to find the best hpam combination (evaluated on the validation dataset) New process TRAIN_FINAL_MODEL: uses the best hpam combination to train the final model and save it #43 Added ProteomicsElasticNet, SingleDrugProteomicsRandomForest to list of known models #38 Reporting all package versions #38 Added UNZIP module for loading and unzipping the drug response datasets instead of handling this in LOAD_RESPONSE: UNZIP_RESPONSE, UNZIP_CS_RESPONSE (for cross-study datasets). #38 Added icon #30 Added the possibility of a leave-tissue-out (LTO) split Changed #53 Changed to large runner for the GitHub Actions because of Docker → Singularity conversion. #42 Moved all publishDir directives to modules.config. #44 Fixed drevalpy versions in conda and docker to 1.3.5: now supporting Python 3.13 #38 Support for AWS: changed the structure of load response and parameter check to conform more to Nextflow best practices. #44 Since drevalpy 1.3.5., the split_early_stopping function is no longer ...