GenerativeProteomics Class

Here you will find the class GenerativeProteomics used during the training and evaluation of the model. The GenerativeProteomics class is responsible for setting up, initializing, and running the GenerativeProteomics imputation process. It reads input arguments, configures the model, trains it, and saves the results.

Methods

  • Main Script (__main__)

    Steps:

    1. Parses Command-Line Arguments using argparse to obtain user-defined settings.

    2. Loads Hyperparameters either from command-line arguments or a JSON configuration file.

    3. Reads the Dataset (CSV/TSV format) and preprocesses it.

    4. Initializes the Generator (G) and Discriminator (D) Networks with a specific architecture.

    5. Creates the Network Class with the model and hyperparameters.

    6. Trains or Evaluates the Model:

      • If a reference dataset is provided, it runs training with reference (train_ref).

      • Otherwise, runs evaluation (evaluate) followed by training (train).

    7. Records Execution Time and stores it in run_time.csv.

    8. Performs Profiling with cProfile to measure execution performance.