Class AutoPeptideML¶

Overview¶

AutoPeptideML is a configurable machine learning workflow class designed for peptide modeling. It integrates data pipelines, representations, model training (with HPO), evaluation, and export.

Class: `AutoPeptideML`¶

Constructor¶

AutoPeptideML(config: dict)

Initializes the AutoPeptideML workflow with a provided configuration dictionary.
Creates output directories and stores pipeline, representation, training, and database settings.

Public Methods¶

`get_pipeline`¶

get_pipeline(pipe_config: Optional[dict] = None) -> Pipeline

Load or construct the preprocessing pipeline.

`get_database`¶

get_database(db_config: Optional[dict] = None) -> Database

Create or load the peptide database with optional negative data support.

`get_reps`¶

get_reps(rep_config: Optional[dict] = None) -> Tuple[Dict[str, RepEngineBase], Dict[str, np.ndarray]]

Load or compute representations for the data.

`get_test`¶

get_test(test_config: Optional[Dict] = None) -> HestiaGenerator

Partition the dataset into training/validation/test using HestiaGenerator.

`get_train`¶

get_train(train_config: Optional[Dict] = None) -> BaseTrainer

Load and return the trainer based on the configuration (supports Optuna and Grid).

`run_hpo`¶

run_hpo() -> Dict

Perform hyperparameter optimization across dataset partitions.

`run_evaluation`¶

run_evaluation(models) -> pd.DataFrame

Run evaluation on the trained models and return a DataFrame of results.

`save_experiment`¶

save_experiment(model_backend: str = 'onnx', save_reps: bool = False, save_test: bool = True, save_all_models: bool = True)

Save the full experiment including models, test partitions, and configuration.

`save_database`¶

save_database()

Export the database to CSV.

`save_models`¶

save_models(ensemble_path: str, backend: str = 'onnx', save_all: bool = True)

Save models using onnx or joblib backends.

`save_reps`¶

save_reps(rep_dir: str)

Save precomputed representations to disk.

`predict`¶

predict(df: pd.DataFrame, feature_field: str, experiment_dir: str, backend: str = 'onnx') -> np.ndarray

Load a saved experiment and predict using the trained ensemble on new data.

Configuration Keys¶

The config dictionary passed to the constructor must include the following keys:

outputdir: str
pipeline: dict or str
representation: dict or str
train: dict or str
databases: dict
test: dict

Dependencies¶

pandas, numpy
yaml, json
hestia
sklearn
skl2onnx, onnxmltools, joblib (optional)

Example Usage¶

from autopipeline.autopeptideml import AutoPeptideML

config = yaml.safe_load(open('config.yml'))
runner = AutoPeptideML(config)
pipeline = runner.get_pipeline()
db = runner.get_database()
reps, x = runner.get_reps()
test = runner.get_test()
trainer = runner.get_train()
models = runner.run_hpo()
evaluation = runner.run_evaluation(models)
runner.save_experiment()

For detailed config templates and supported options, see the corresponding YAML schema documentation.

Class AutoPeptideML¶

Overview¶