seqgra.seqgra module

MIT - CSAIL - Gifford Lab - seqgra

seqgra complete pipeline: 1. generate data based on data definition (once), see run_simulator.py 2. train model on data (once), see run_learner.py 3. evaluate model performance with SIS, see run_sis.py

@author: Konstantin Krismer

create_parser()

main()

run_seqgra(data_def_file: Optional[str], data_folder: Optional[str], model_def_file: Optional[str], evaluator_ids: Optional[List[str]], output_dir: str, in_memory: bool, print_info: bool, silent: bool, remove_existing_data: bool, gpu_id: int, no_checks: bool, eval_sets: Optional[List[str]], eval_n: Optional[int], eval_n_per_label: Optional[int], eval_suppress_plots: Optional[bool], eval_fi_predict_threshold: Optional[float], eval_sis_predict_threshold: Optional[float], eval_grad_importance_threshold: Optional[float]) → None