7.5. Pickaxe

Pickaxe.py: Create network expansions from reaction rules and compounds.

This module generates new compounds from user-specified starting compounds using a set of SMARTS-based reaction rules.

class minedatabase.pickaxe.Pickaxe(rule_list: Optional[str] = None, coreactant_list: Optional[str] = None, explicit_h: bool = False, kekulize: bool = False, neutralise: bool = True, errors: bool = True, inchikey_blocks_for_cid: int = 1, database: Optional[str] = None, database_overwrite: bool = False, mongo_uri: bool = 'mongodb://localhost:27017', image_dir: Optional[str] = None, quiet: bool = True, react_targets: bool = True, filter_after_final_gen: bool = True, prune_between_gens: bool = False)

Class to generate expansions with compounds and reaction rules.

This class generates new compounds from user-specified starting compounds using a set of SMARTS-based reaction rules. It may be initialized with a text file containing the reaction rules and coreactants or this may be done on an ad hoc basis.

Parameters

rule_liststr: Filepath of rules.
coreactant_liststr: Filepath of coreactants.
explicit_hbool, optional: Whether rules utilize explicit hydrogens, by default True.
kekulizebool, optional: Whether or not to kekulize compounds before reaction, by default False.
neutralisebool, optional: Whether or not to neutralise compounds, by default True.
errorsbool, optional: Whether or not to print errors to stdout, by default True.
inchikey_blocks_for_cidint, optional: How many blocks of the InChI key to use for the compound id, by default 1.
databasestr, optional: Name of the database where to save results, by default None.
database_overwritebool, optional: Whether or not to erase existing database in event of a collision, by default False.
mongo_uribool, optional: uri for the mongo client, by default ‘mongodb://localhost:27017’.
image_dirstr, optional: Filepath where images should be saved, by default None.
quietbool, optional: Whether to silence warnings, by default False.
react_targetsbool, optional: Whether or not to apply reactions to generated compounds that match targets, by default True.
filter_after_final_genbool, optional: Whether to apply filters after final expansion, by default True.
prune_between_gensbool, optional: Whether to prune network between generations if using filters

Attributes

operators: dict: Reaction operators to transform compounds with.
coreactants: dict: Coreactants required by the operators.
compounds: dict: Compounds in the pickaxe network.
reactions: dict: Reactions in the pickaxe network.
generation: int: The current generation
explicit_hbool: Whether rules utilize explicit hydrogens.
kekulizebool: Whether or not to kekulize compounds before reaction.
neutralisebool: Whether or not to neutralise compounds.
fragmented_molsbool: Whether or not to allow fragmented molecules.
radical_checkbool: Whether or not to check and remove radicals.
image_dirstr, optional: Filepath where images should be saved.
errorsbool: Whether or not to print errors to stdout.
quietbool: Whether or not to silence warnings.
filters: List[object]: A list of filters to apply during the expansion.
targetsdict: Molecules to be targeted during expansions.
target_smiles: List[str]: The SMILES of all the targets.
react_targetsbool: Whether or not to react targets when generated.
filter_after_final_genbool: Whether or not to filter after the last expansion.
prune_between_gensbool, optional: Whether to prune network between generations if using filters.
mongo_uristr: The connection string to the mongo database.
cid_num_inchi_blocksint: How many blocks of the inchi-blocks to use to generate the compound id.

assign_ids() → None

Assign a numerical ID to compounds (and reactions).

Assign IDs that are unique only to the CURRENT run.

find_minimal_set(white_list: Set[str]) → Tuple[set, set]

Find the minimal set of compounds and reactions given a white list.

Given a whitelist this function finds the minimal set of compound and reactions ids that comprise the set.

Parameters

white_listSet[str]: List of compound_ids to use to filter reaction network to.

Returns

Tuple[set, set]: The filtered compounds and reactions.

load_compound_set(compound_file: Optional[str] = None, id_field: str = 'id') → str

Load compounds for expansion into pickaxe.

Parameters

compound_filestr, optional: Filepath of compounds, by default None.
id_fieldstr, optional: Header value of compound id in input file, by default ‘id’.

Returns

str: List of SMILES that were succesfully loaded into pickaxe.

Raises

ValueError: No file specified for loading.

load_pickled_pickaxe(fname: str) → None

Load pickaxe from pickle.

Load pickled pickaxe object.

Parameters

fnamestr: filename to read (must be .pk).

load_targets(target_compound_file: Optional[str], id_field: str = 'id') → None

Load targets into pickaxe.

Parameters

target_compound_filestr: Filepath of target compounds.
id_fieldstr, optional: Header value of compound id in input file, by default ‘id’.

pickle_pickaxe(fname: str) → None

Pickle key pickaxe items.

Pickle pickaxe object to be loaded in later.

Parameters

fnamestr: filename to save (must be .pk).

prune_network(white_list: list, print_output: str = True) → None

Prune the reaction network to a list of targets.

Prune the predicted reaction network to only compounds and reactions that terminate in a specified white list of compounds.

Parameters

white_listlist: A list of compound ids to filter the network to.
print_outputbool: Whether or not to print output

prune_network_to_targets() → None

Prune the reaction network to the target compounds.

Prune the predicted reaction network to only compounds and reactions that terminate in the target compounds.

save_to_SBML(file_name: str, save_reactions_uniprot: bool = False, uniprot_save_style: str = 'grouped') → None

Save pickaxe run to an SBML file.

This function saves the species and reactions to an SBML file with annotations. Specifically, the species will be annotated with their SMILES and reactions annotated with their operator and uniprot ids (if available and desired).

Parameters

file_namestr

The file name to save the SBML at.

save_reactions_uniprotbool

Whether or not to save the uniprot ids for a reaction operator (if available)

uniprot_save_stylestr

The stle to save uniprot ids in. There are two options:: grouped : all uniprot information is stored in a semicolon delimited list individual : uniprot information is saved individually as a link to the uniprot website

save_to_mine(processes: int = 1, indexing: bool = True, write_core: bool = False) → None

Save pickaxe run to MINE database.

Parameters

processesint, optional: Number of processes to use, by default 1.
indexingbool, optional: Whether or not to add indexes, by default True.
write_corebool, optional: Whether or not to write to core database, by default False.

transform_all(processes: int = 1, generations: int = 1) → None

Transform compounds with reaction operators.

Apply reaction rules to compounds and generate a specified number of new generations.

Parameters

processesint, optional: Number of processes to run in parallel, by default 1.
generationsint, optional: Number of generations to create, by default 1.

write_compound_output_file(path: str, dialect: str = 'excel-tab') → None

Write compounds to an output file.

Parameters

pathstr: Path to write data.
dialectstr, optional: Dialect of the output, by default ‘excel-tab’.

write_reaction_output_file(path: str, delimiter: str = '\t') → None

Write all reaction data to the specified path.

Parameters

pathstr: Path to write data.
delimiterstr, optional: Delimiter for the output file, by default ‘t’.