config.yaml

# config.yaml

path:
  # Path of the clustering result from MS-Cluster
  cluster_path: ./data/external/mscluster

  # Path of a folder containing MS/MS spectra (MGF format)
  mgf_path: ./data/raw

  search_results:
    # Path and filename for de_novo search results (CSV)
    de_novo:
      path: ./data/external/denovo
      filename: PXD014222_Casanovo.csv

    # Path and filename for database search results (CSV)
    # Used only in the training process.
    db:
      path: ./data/external/db
      filename: PXD014222_db.csv

  # Path to save project outputs
  save_path: ./data/interim

  # Path of a folder containing MS/MS denoised spectra (MGF format)
  denoised_mgf_path: ./data/processed/denoised_mgf

  # Path of a folder containing MS/MS spectra for XCorr calculation (MGF format)
  xcorr_mgf_path: ./data/processed/xcorr_mgf

  # Filename for the interim report (Feature file: CSV)
  interim_report: dataset_top2.csv

  # Path of the XCorr calculation results from CometX
  xcorr_results_path: ./data/external/xcorr

  # Path where the trained model will be saved
  model_save:
    path: ./models
    filename: novnorank.h5 # (h5)

  # Path to the pretrained model for inference
  pretrained_model:
    path: ./pretrained
    filename: NovoRank_Casanovo.h5  # Example (h5)

  # Filename for the final report (CSV)
  final_report: results_top1.csv

params:
  # Whether to train the model (True: training process, False: Inference process)
  train: True

  # Whether to create a test dataset (True: create a test dataset, False: do not create)
  # Used only in the training process.
  test_set: True

  # Whether to save the model checkpoint during training (True: save checkpoint every epoch, False: do not save)
  # Used only in the training process.
  checkpoint: True

  # Whether to report the top 10 results (True: report top 10 candidate results, False: do not report)
  # Used only in the inference process.
  top_10: False

  # fine tuning ?