Function peptide_identifications_to_df

src/pyOpenMS/pyopenms/_dataframes.py:485–582

Converts a list of peptide identifications to a pandas DataFrame. Parameters: peps (List[PeptideIdentification]): list of PeptideIdentification objects decode_ontology (bool): decode meta value names default_missing_values: default value for missing values for each data type expo

(peps: List[_PeptideIdentification], decode_ontology : bool = True,
                                  default_missing_values: dict = {bool: False, int: -9999, float: _np.nan, str: ''},
                                  export_unidentified : bool = True)

# TODO think about the best way for such top-level function. IMHO in python, encapsulation in a stateless class in unnecessary.
#   We should probably not just import this whole submodule without prefix.
def peptide_identifications_to_df(peps: List[_PeptideIdentification], decode_ontology : bool = True,
                                  default_missing_values: dict = {bool: False, int: -9999, float: _np.nan, str: ''},
                                  export_unidentified : bool = True):
    """Converts a list of peptide identifications to a pandas DataFrame.
    Parameters:
    peps (List[PeptideIdentification]): list of PeptideIdentification objects
    decode_ontology (bool): decode meta value names
    default_missing_values: default value for missing values for each data type
    export_unidentified: export PeptideIdentifications without PeptideHit
    Returns:
    pandas.DataFrame: peptide identifications in a DataFrame
    """
    switchDict = {bool: '?', int: 'i', float: 'f', str: 'U100'}

    # filter out PeptideIdentifications without PeptideHits if export_unidentified == False
    count = len(peps)
    if not export_unidentified:
        count = sum(len(pep.getHits()) > 0 for pep in peps)

    # get all possible metavalues
    metavals = []
    types = []
    mainscorename = "score"
    for pep in peps:
        hits = pep.getHits()
        if not len(hits) == 0:
            mvs = []
            hits[0].getKeys(mvs)
            metavals += mvs
            mainscorename = pep.getScoreType()

    metavals = list(set(metavals))

    # get type of all metavalues
    for k in metavals:
        if k == b"target_decoy":
            types.append('?')
        else:
            for p in peps:
                hits = p.getHits()
                if not len(hits) == 0:
                    mv = hits[0].getMetaValue(k)
                    types.append(switchDict[type(mv)])
                    break

    # get default value for each type in types to append if there are no hits in a PeptideIdentification
    def get_key(val):
        for key, value in switchDict.items():
            if val == value:
                return key
    dmv = [default_missing_values[get_key(t)] for t in types]

    decodedMVs = [m.decode("utf-8") for m in metavals]
    if decode_ontology:
        cv = _ControlledVocabulary()
        cv.loadFromOBO("psims", _File.getOpenMSDataPath() + "/CV/psi-ms.obo")
        clearMVs = [cv.getTerm(m).name if m.startswith("MS:") else m for m in decodedMVs]
    else: