Module deeporigin.src.utilities.props

Functions

def predict_properties(smiles: str) ‑> Dict[str, Any] | None

Expand source code

def predict_properties(smiles: str,) -> Optional[Dict[str, Any]]:
    """Predicts chemical properties for a given SMILES string.

    This function sends a request to predict various chemical properties including hERG,
    logD, logP, logS, ames test results, and cytochrome P450 interactions for a given
    chemical compound represented by its SMILES notation.

    Args:
        smiles (str): A SMILES (Simplified Molecular Input Line Entry System) string
            representing the chemical compound.

    Returns:
        Optional[Dict[str, Any]]: A dictionary containing the predicted properties if successful,
            with property names as keys and their predicted values as values.
            Returns None if the prediction fails for any reason (network error, invalid response, etc.).

    Example:
        >>> properties = predict_properties("CC(=O)OC1=CC=CC=C1C(=O)O")
        >>> if properties:
        ...     print(properties["logP"])

    Raises:
        No explicit raises, but handles various exceptions internally and returns None on failure.

    Note:
        The function logs its progress and any errors through DEFAULT_LOGGER.
    """
    DEFAULT_LOGGER.log_info("Starting property prediction process.")

    properties_to_predict = {
        "hERG": True,
        "logD": True,
        "logP": True,
        "logS": True,
        "ames": True,
        "cyp": True
    }
    DEFAULT_LOGGER.log_info(f"Properties to predict: {list(properties_to_predict.keys())}")

    request_body = {
        "smiles": smiles,
        "predict": properties_to_predict
    }
    DEFAULT_LOGGER.log_info("Prepared request body for property prediction.")

    client = Client()
    try:
        DEFAULT_LOGGER.log_info("Sending property prediction request.")
        response = client.post_request(
            endpoint="properties",
            logger=DEFAULT_LOGGER,
            data=request_body
        )
        DEFAULT_LOGGER.log_info("Received response from property prediction request.")
    except Exception as e:
        DEFAULT_LOGGER.log_error(f"Failed to send property prediction request: {e}")
        return None

    if response.status_code == 200:
        try:
            response_data = response.json()
            DEFAULT_LOGGER.log_info("Property prediction successful.")
            return response_data
        except ValueError:
            DEFAULT_LOGGER.log_error("Failed to parse JSON response.")
            return None
    else:
        try:
            error_message = response.json().get("msg", "Unknown error occurred.")
            DEFAULT_LOGGER.log_error(f"Property prediction failed: {error_message}")
        except ValueError:
            DEFAULT_LOGGER.log_error("Failed to parse error message from response.")
        return None

Predicts chemical properties for a given SMILES string.

This function sends a request to predict various chemical properties including hERG, logD, logP, logS, ames test results, and cytochrome P450 interactions for a given chemical compound represented by its SMILES notation.

Args

smiles : str

A SMILES (Simplified Molecular Input Line Entry System) string representing the chemical compound.

Returns

Optional[Dict[str, Any]]

A dictionary containing the predicted properties if successful, with property names as keys and their predicted values as values. Returns None if the prediction fails for any reason (network error, invalid response, etc.).

Example

>>> properties = predict_properties("CC(=O)OC1=CC=CC=C1C(=O)O")
>>> if properties:
...     print(properties["logP"])

Raises

No explicit raises, but handles various exceptions internally and returns None on failure.

Note

The function logs its progress and any errors through DEFAULT_LOGGER.

def protonate(smiles: str, pH: float = 7.4, filter_percentage: float = 1)

Expand source code

def protonate(smiles: str, pH: float=7.4, filter_percentage: float=1):
    """Predicts the protonation state of a molecule at a given pH.
    This function takes a SMILES string representation of a molecule and predicts its
    protonation state at the specified pH using a molecular properties client.
    
    Args:
        smiles (str): SMILES string representation of the input molecule
        pH (float, optional): The pH value at which to predict protonation. Defaults to 7.4
        filter_percentage (float, optional): Filter threshold for protonation prediction. 
            Defaults to 1.0
    
    Returns:
        str or None: SMILES string of the protonated molecule if successful, None if failed
    
    Raises:
        Exception: If the protonation prediction request fails
    
    Example:
        >>> protonated_smiles = protonate("CC(=O)O", pH=7.4)
    """
    from deeporigin.src.properties import MolecularPropertiesClient
    
    try:
        client = MolecularPropertiesClient()
        DEFAULT_LOGGER.log_info("Protonation prediction process started.")

        protonation_output = client.protonate(
            entry=[smiles],
            pH=pH,
            filter_percentage=filter_percentage,
            html_output=False
        )[0]
        smiles = protonation_output["protonation"]["smiles_list"][0]
        
        return smiles
    except Exception as e:
        DEFAULT_LOGGER.log_error(f"Failed to send property prediction request: {e}")
        return None

Predicts the protonation state of a molecule at a given pH. This function takes a SMILES string representation of a molecule and predicts its protonation state at the specified pH using a molecular properties client.

Args

smiles : str

SMILES string representation of the input molecule

pH : float, optional

The pH value at which to predict protonation. Defaults to 7.4

filter_percentage : float, optional

Filter threshold for protonation prediction. Defaults to 1.0

Returns

str or None

SMILES string of the protonated molecule if successful, None if failed

Raises

Exception

If the protonation prediction request fails

Example

>>> protonated_smiles = protonate("CC(=O)O", pH=7.4)