comparison COBRAxy/rps_generator.py @ 381:0a3ca20848f3 draft

Uploaded

380:03a7ba63813f

        str : the internal :str unique identifier of that metabolite, used in all other parts of the model in use.
        An empty string is returned if a match isn't found.
    """
    name = clean_metabolite_name(name)
    for id, synonyms in syn_dict.items():
        if name in synonyms: return id
    
    return ""

############################ check_missing_metab ####################################
def check_missing_metab(reactions: Dict[str, Dict[str, int]], dataset_by_rows: Dict[str, List[float]], cell_lines_amt :int) -> List[str]:
    """
    Check for missing metabolites in the abundances dictionary compared to the reactions dictionary and update abundances accordingly.

    Parameters:
        reactions (dict): A dictionary representing reactions where keys are reaction names and values are dictionaries containing metabolite names as keys and stoichiometric coefficients as values.
        dataset_by_rows (dict): A dictionary representing abundances where keys are metabolite names and values are their corresponding abundances for all cell lines.
        cell_lines_amt : amount of cell lines, needed to add a new list of abundances for missing metabolites.

    Returns:
        list[str] : list of metabolite names that were missing in the original abundances dictionary and thus their aboundances were set to 1.

        substrateFreqTable (dict): A dictionary where each metabolite name (key) is associated with how many times it shows up in the model's reactions (value).

    Returns:
        None
    """
    cell_lines = dataset[0][1:]
    abundances_dict = {}

    translationIsApplied = ARGS.reaction_choice == "default"
    for row in dataset[1:]:
        id = get_metabolite_id(row[0], syn_dict) if translationIsApplied else row[0]
        if id: abundances_dict[id] = list(map(utils.Float(), row[1:]))
    
    missing_list = check_missing_metab(reactions, abundances_dict, len((cell_lines)))
    
    rps_scores :Dict[Dict[str, float]] = {}
    for pos, cell_line_name in enumerate(cell_lines):
        abundances = { metab : abundances[pos] for metab, abundances in abundances_dict.items() }
        rps_scores[cell_line_name] = calculate_rps(reactions, abundances, black_list, missing_list, substrateFreqTable)
    
    df = pd.DataFrame.from_dict(rps_scores)
    
    df.index.name = 'Reactions'
    df.to_csv(ARGS.rps_output, sep='\t', na_rep='None', index=True)

############################ main ####################################
def main(args:List[str] = None) -> None:

    with open(ARGS.tool_dir + '/local/pickle files/synonyms.pickle', 'rb') as sd:
        syn_dict = pk.load(sd)

    dataset = utils.readCsv(utils.FilePath.fromStrPath(ARGS.input), '\t', skipHeader = False)

    if ARGS.reaction_choice == 'default':
        reactions = pk.load(open(ARGS.tool_dir + '/local/pickle files/reactions.pickle', 'rb'))
        substrateFreqTable = pk.load(open(ARGS.tool_dir + '/local/pickle files/substrate_frequencies.pickle', 'rb'))
    
    elif ARGS.reaction_choice == 'custom':
        reactions = reactionUtils.parse_custom_reactions(ARGS.custom)
        substrateFreqTable = {}
        for _, substrates in reactions.items():
            for substrateName, _ in substrates.items():
                if substrateName not in substrateFreqTable: substrateFreqTable[substrateName] = 0
                substrateFreqTable[substrateName] += 1

    rps_for_cell_lines(dataset, reactions, black_list, syn_dict, substrateFreqTable)
    print('Execution succeded')

##############################################################################
if __name__ == "__main__": main()

381:0a3ca20848f3

        str : the internal :str unique identifier of that metabolite, used in all other parts of the model in use.
        An empty string is returned if a match isn't found.
    """
    name = clean_metabolite_name(name)
    for id, synonyms in syn_dict.items():
        if name in synonyms:
            return id
    
    return ""

############################ check_missing_metab ####################################
def check_missing_metab(reactions: Dict[str, Dict[str, int]], dataset_by_rows: Dict[str, List[float]], cell_lines_amt :int) -> List[str]:
    """
    Check for missing metabolites in the abundances dictionary compared to the reactions dictionary and update abundances accordingly.

    Parameters:
        reactions (dict): A dictionary representing reactions where keys are reaction names and values are dictionaries containing metabolite names as keys and 
                          stoichiometric coefficients as values.
        dataset_by_rows (dict): A dictionary representing abundances where keys are metabolite names and values are their corresponding abundances for all cell lines.
        cell_lines_amt : amount of cell lines, needed to add a new list of abundances for missing metabolites.

    Returns:
        list[str] : list of metabolite names that were missing in the original abundances dictionary and thus their aboundances were set to 1.

        substrateFreqTable (dict): A dictionary where each metabolite name (key) is associated with how many times it shows up in the model's reactions (value).

    Returns:
        None
    """

    cell_lines = dataset[0][1:]
    abundances_dict = {}

    translationIsApplied = ARGS.reaction_choice == "default"
    for row in dataset[1:]:
        id = get_metabolite_id(row[0], syn_dict) #if translationIsApplied else row[0]
        if id: 
            abundances_dict[id] = list(map(utils.Float(), row[1:]))

    missing_list = check_missing_metab(reactions, abundances_dict, len((cell_lines)))

    rps_scores :Dict[Dict[str, float]] = {}
    for pos, cell_line_name in enumerate(cell_lines):
        abundances = { metab : abundances[pos] for metab, abundances in abundances_dict.items() }

        rps_scores[cell_line_name] = calculate_rps(reactions, abundances, black_list, missing_list, substrateFreqTable)
    
    df = pd.DataFrame.from_dict(rps_scores)
    df = df.loc[list(reactions.keys()),:]
    print(df.head(10))
    df.index.name = 'Reactions'
    df.to_csv(ARGS.rps_output, sep='\t', na_rep='None', index=True)

############################ main ####################################
def main(args:List[str] = None) -> None:

    with open(ARGS.tool_dir + '/local/pickle files/synonyms.pickle', 'rb') as sd:
        syn_dict = pk.load(sd)

    dataset = utils.readCsv(utils.FilePath.fromStrPath(ARGS.input), '\t', skipHeader = False)
    tmp_dict = None
    if ARGS.reaction_choice == 'default':
        reactions = pk.load(open(ARGS.tool_dir + '/local/pickle files/reactions.pickle', 'rb'))
        substrateFreqTable = pk.load(open(ARGS.tool_dir + '/local/pickle files/substrate_frequencies.pickle', 'rb'))
    
    elif ARGS.reaction_choice == 'custom':
        reactions = reactionUtils.parse_custom_reactions(ARGS.custom)
        for r, s in reactions.items():
            tmp_list = list(s.keys())
            for k in tmp_list:
                if k[-2] == '_':
                    s[k[:-2]] = s.pop(k)
        substrateFreqTable = {}
        for _, substrates in reactions.items():
            for substrateName, _ in substrates.items():
                if substrateName not in substrateFreqTable: substrateFreqTable[substrateName] = 0
                substrateFreqTable[substrateName] += 1

        print(f"Reactions: {reactions}")
        print(f"Substrate Frequencies: {substrateFreqTable}")
        print(f"Synonyms: {syn_dict}")
        tmp_dict = {}
        for metabName, freq in substrateFreqTable.items():
            tmp_metabName = clean_metabolite_name(metabName)
            for syn_key, syn_list in syn_dict.items():
                if tmp_metabName in syn_list or tmp_metabName == clean_metabolite_name(syn_key):
                    print(f"Mapping {tmp_metabName} to {syn_key}")
                    tmp_dict[syn_key] = syn_list
                    tmp_dict[syn_key].append(tmp_metabName)

    rps_for_cell_lines(dataset, reactions, black_list, syn_dict, substrateFreqTable)
    print('Execution succeded')

##############################################################################
if __name__ == "__main__": main()