rxnutils.pipeline.actions package

Submodules

rxnutils.pipeline.actions.dataframe_mod module

Module containing actions that modify the dataframe in some way

class rxnutils.pipeline.actions.dataframe_mod.DropColumns(columns)

Bases: object

Drops columns specified in ‘columns’

yaml example:

drop_columns:

columns:

• NRingChange

• RingBondMade

Parameters:

columns (list[str])

pretty_name: ClassVar[str] = 'drop_columns'

columns: list[str]

class rxnutils.pipeline.actions.dataframe_mod.DropDuplicates(key_columns)

Bases: object

Action for dropping duplicates from dataframe

Parameters:

key_columns (list[str])

pretty_name: ClassVar[str] = 'drop_duplicates'

key_columns: list[str]

class rxnutils.pipeline.actions.dataframe_mod.DropRows(indicator_columns)

Bases: object

Drops rows according to boolean in ‘indicator_columns’. True => Keep, False => Drop.

yaml example:

drop_rows:

indicator_columns:

• is_sanitizable

• is_sanitizable2

Parameters:

indicator_columns (list[str])

pretty_name: ClassVar[str] = 'drop_rows'

indicator_columns: list[str]

class rxnutils.pipeline.actions.dataframe_mod.KeepColumns(columns)

Bases: object

Drops columns not specified in ‘columns’

yaml example:

keep_columns:

columns:

• id

• classification

• rsmi_processed

Parameters:

columns (list[str])

pretty_name: ClassVar[str] = 'keep_columns'

columns: list[str]

class rxnutils.pipeline.actions.dataframe_mod.RenameColumns(in_columns, out_columns)

Bases: object

Renames columns specified in ‘in_columns’ to the names specified in ‘out_columns’

yaml example:

rename_columns:

in_columns:

• column1

• column2

out_columns:

• column1_renamed

• column2_renamed

Parameters:

• in_columns (list[str])

• out_columns (list[str])

pretty_name: ClassVar[str] = 'rename_columns'

in_columns: list[str]

out_columns: list[str]

class rxnutils.pipeline.actions.dataframe_mod.QueryDataframe(query)

Bases: object

Uses dataframe query to produce a new (smaller) dataframe. Query must conform to pandas.query()

yaml file example: Keeping only rows where the has_stereo columns is True:

query_dataframe:

query: has_stereo == True

Parameters:

query (str)

pretty_name: ClassVar[str] = 'query_dataframe'

query: str

class rxnutils.pipeline.actions.dataframe_mod.StackColumns(in_columns, out_column='StackedColumns')

Bases: object

Stacks the specified in_columns under a new column name (out_column), multiplies the rest of the columns as appropriate

yaml control file example:

stack_columns:

in_columns:

• rsmi_processed

• rsmi_inverted_stereo

out_column: rsmi_processed

Parameters:

• in_columns (list[str])

• out_column (str)

pretty_name: ClassVar[str] = 'stack_columns'

in_columns: list[str]

out_column: str = 'StackedColumns'

class rxnutils.pipeline.actions.dataframe_mod.StackMultiColumns(stack_columns, target_columns)

Bases: object

Stacks the specified target_columns on top of the stack_columns after renaming.

Example Yaml:

stack_multi_columns:

stack_columns:

• rsmi_inverted_stereo

• PseudoHash_inverted_stereo

target_columns:

• rsmi_processed

• PseudoHash

Parameters:

• stack_columns (list[str])

• target_columns (list[str])

pretty_name: ClassVar[str] = 'stack_multi_columns'

stack_columns: list[str]

target_columns: list[str]

rxnutils.pipeline.actions.reaction_mod module

Module containing actions on reactions that modify the reaction in some way

class rxnutils.pipeline.actions.reaction_mod.DesaltMolecules(in_column, out_column='RxnDesalted', keep_something=False)

Bases: ReactionActionMixIn

Action for desalting molecules

Parameters:

• in_column (str)

• out_column (str)

• keep_something (bool)

pretty_name: ClassVar[str] = 'desalt_molecules'

in_column: str

out_column: str = 'RxnDesalted'

keep_something: bool = False

class rxnutils.pipeline.actions.reaction_mod.NameRxn(in_column, options='', nm_rxn_column='NextMoveRxnSmiles', nmc_column='NMC')

Bases: object

Action for calling namrxn

Parameters:

• in_column (str)

• options (str)

• nm_rxn_column (str)

• nmc_column (str)

pretty_name: ClassVar[str] = 'namerxn'

in_column: str

options: str = ''

nm_rxn_column: str = 'NextMoveRxnSmiles'

nmc_column: str = 'NMC'

class rxnutils.pipeline.actions.reaction_mod.NeutralizeMolecules(in_column, out_column='RxnNeutralized')

Bases: ReactionActionMixIn

Action for neutralizing molecules

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'neutralize_molecules'

in_column: str

out_column: str = 'RxnNeutralized'

class rxnutils.pipeline.actions.reaction_mod.ReactantsToReagents(in_column, out_column='RxnSmilesWithTrueReagents')

Bases: ReactionActionMixIn

Action for converting reactants to reagents

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'reactants2reagents'

in_column: str

out_column: str = 'RxnSmilesWithTrueReagents'

class rxnutils.pipeline.actions.reaction_mod.ReagentsToReactants(in_column, out_column='RxnSmilesAllReactants')

Bases: object

Action for converting reagents to reactants

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'reagents2reactants'

in_column: str

out_column: str = 'RxnSmilesAllReactants'

class rxnutils.pipeline.actions.reaction_mod.RemoveAtomMapping(in_column, out_column='RxnSmilesNoAtomMap')

Bases: ReactionActionMixIn

Action for removing all atom mapping

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'remove_atom_mapping'

in_column: str

out_column: str = 'RxnSmilesNoAtomMap'

class rxnutils.pipeline.actions.reaction_mod.RemoveStereoInfo(in_column, out_column='RxnSmilesNoStereo')

Bases: object

Action for removing stero information

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'remove_stereo_info'

in_column: str

out_column: str = 'RxnSmilesNoStereo'

class rxnutils.pipeline.actions.reaction_mod.InvertStereo(in_column, out_column='RxnSmilesInvertedStereo')

Bases: object

Action for inverting stero information

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'invert_stereo'

in_column: str

out_column: str = 'RxnSmilesInvertedStereo'

class rxnutils.pipeline.actions.reaction_mod.IsotopeInfo(in_column, isotope_column='Isotope', out_column='RxnSmilesWithoutIsotopes', match_regex='\\[(?P<mass>[0-9]+)(?P<symbol>[A-Za-z][a-z]*)(?P<spec>[^\\]]+)*\\]', sub_regex='[\\g<symbol>\\g<spec>]')

Bases: object

Action creating and modifying isotope information

Parameters:

• in_column (str)

• isotope_column (str)

• out_column (str)

• match_regex (str)

• sub_regex (str)

pretty_name: ClassVar[str] = 'isotope_info'

in_column: str

isotope_column: str = 'Isotope'

out_column: str = 'RxnSmilesWithoutIsotopes'

match_regex: str = '\\[(?P<mass>[0-9]+)(?P<symbol>[A-Za-z][a-z]*)(?P<spec>[^\\]]+)*\\]'

sub_regex: str = '[\\g<symbol>\\g<spec>]'

class rxnutils.pipeline.actions.reaction_mod.RemoveExtraAtomMapping(in_column, out_column='RxnSmilesReassignedAtomMap')

Bases: ReactionActionMixIn

Action for removing extra atom mapping

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'remove_extra_atom_mapping'

in_column: str

out_column: str = 'RxnSmilesReassignedAtomMap'

class rxnutils.pipeline.actions.reaction_mod.RemoveUnchangedProducts(in_column, out_column='RxnNoUnchangedProd')

Bases: ReactionActionMixIn

Compares the products with the reagents and reactants and remove unchanged products.

Protonation is considered a difference, As example, if there’s a HCl in the reagents and a Cl- in the products, it will not be removed.

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'remove_unchanged_products'

in_column: str

out_column: str = 'RxnNoUnchangedProd'

class rxnutils.pipeline.actions.reaction_mod.RemoveUnsanitizable(in_column, out_column='RxnSanitizable', bad_column='BadMolecules')

Bases: ReactionActionMixIn

Action for removing unsanitizable reactions

Parameters:

• in_column (str)

• out_column (str)

• bad_column (str)

pretty_name: ClassVar[str] = 'remove_unsanitizable'

in_column: str

out_column: str = 'RxnSanitizable'

bad_column: str = 'BadMolecules'

class rxnutils.pipeline.actions.reaction_mod.RDKitRxnRoles(in_column, out_column='RxnRoleAssigned')

Bases: object

Action for assigning roles based on RDKit algorithm

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'rdkit_RxnRoleAssignment'

in_column: str

out_column: str = 'RxnRoleAssigned'

class rxnutils.pipeline.actions.reaction_mod.RxnMapper(in_column, out_column='RxnmapperRxnSmiles', rxnmapper_command='conda run -p ${RXNMAPPER_ENV_PATH} python /home/runner/work/reaction_utils/reaction_utils/rxnutils/data/mapping.py')

Bases: object

Action for mapping reactions with the RXNMapper tool

Parameters:

• in_column (str)

• out_column (str)

• rxnmapper_command (str)

pretty_name: ClassVar[str] = 'rxnmapper'

in_column: str

out_column: str = 'RxnmapperRxnSmiles'

rxnmapper_command: str = 'conda run -p ${RXNMAPPER_ENV_PATH} python /home/runner/work/reaction_utils/reaction_utils/rxnutils/data/mapping.py'

class rxnutils.pipeline.actions.reaction_mod.SplitReaction(in_column, out_columns)

Bases: ReactionActionMixIn

Action for splitting reaction into components

Parameters:

• in_column (str)

• out_columns (list[str])

pretty_name: ClassVar[str] = 'split_reaction'

in_column: str

out_columns: list[str]

class rxnutils.pipeline.actions.reaction_mod.AtomMapTagDisconnectionSite(in_column='RxnSmilesClean', out_column='products_atom_map_tagged')

Bases: ReactionActionMixIn

Action for tagging disconnection site in products with atom-map ‘[<atom>:1]’.

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'atom_map_tag_disconnection_site'

in_column: str = 'RxnSmilesClean'

out_column: str = 'products_atom_map_tagged'

class rxnutils.pipeline.actions.reaction_mod.ConvertAtomMapDisconnectionTag(in_column='products_atom_map_tagged', out_column_tagged='products_tagged', out_column_reconstructed='products_reconstructed')

Bases: ReactionActionMixIn

Action for converting atom-map tagging to exclamation mark tagging.

yaml example:

convert_atom_map_disconnection_tag:

in_column_tagged: products_atom_map_tagged in_column_untagged: products out_column_tagged: products_tagged out_column_reconstructed: products_reconstructed

Parameters:

• in_column (str)

• out_column_tagged (str)

• out_column_reconstructed (str)

pretty_name: ClassVar[str] = 'convert_atom_map_disconnection_tag'

in_column: str = 'products_atom_map_tagged'

out_column_tagged: str = 'products_tagged'

out_column_reconstructed: str = 'products_reconstructed'

class rxnutils.pipeline.actions.reaction_mod.TrimRxnSmiles(in_column, out_column='RxnSmiles', smiles_column_index=0)

Bases: object

Action from trimming reaction SMILES

Parameters:

• in_column (str)

• out_column (str)

• smiles_column_index (int)

pretty_name: ClassVar[str] = 'trim_rxn_smiles'

in_column: str

out_column: str = 'RxnSmiles'

smiles_column_index: int = 0

rxnutils.pipeline.actions.reaction_props module

Module containing actions on reactions that doesn’t modify the reactions only compute properties of them

class rxnutils.pipeline.actions.reaction_props.CountComponents(in_column, nreactants_column='NReactants', nmapped_reactants_column='NMappedReactants', nreagents_column='NReagents', nmapped_reagents_column='NMappedReagents', nproducts_column='NProducts', nmapped_products_column='NMappedProducts')

Bases: ReactionActionMixIn

Action for counting reaction components

Parameters:

• in_column (str)

• nreactants_column (str)

• nmapped_reactants_column (str)

• nreagents_column (str)

• nmapped_reagents_column (str)

• nproducts_column (str)

• nmapped_products_column (str)

pretty_name: ClassVar[str] = 'count_components'

in_column: str

nreactants_column: str = 'NReactants'

nmapped_reactants_column: str = 'NMappedReactants'

nreagents_column: str = 'NReagents'

nmapped_reagents_column: str = 'NMappedReagents'

nproducts_column: str = 'NProducts'

nmapped_products_column: str = 'NMappedProducts'

class rxnutils.pipeline.actions.reaction_props.CountElements(in_column, out_column='ElementCount')

Bases: object

Action for counting elements in reactants

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'count_elements'

in_column: str

out_column: str = 'ElementCount'

class rxnutils.pipeline.actions.reaction_props.HasStereoInfo(in_column, out_column='HasStereo')

Bases: object

Action for checking stereo info

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'has_stereo_info'

in_column: str

out_column: str = 'HasStereo'

class rxnutils.pipeline.actions.reaction_props.HasUnmappedRadicalAtom(in_column, out_column='HasUnmappedRadicalAtom')

Bases: object

Action for flagging if reaction has any unmapped radical atoms

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'hasunmappedradicalatom'

in_column: str

out_column: str = 'HasUnmappedRadicalAtom'

class rxnutils.pipeline.actions.reaction_props.HasUnsanitizableReactants(rsmi_column, bad_columns, out_column='HasUnsanitizableReactants')

Bases: object

Action for flagging if reaction has any unsanitizable reactants

Parameters:

• rsmi_column (str)

• bad_columns (List[str])

• out_column (str)

pretty_name: ClassVar[str] = 'unsanitizablereactants'

rsmi_column: str

bad_columns: List[str]

out_column: str = 'HasUnsanitizableReactants'

class rxnutils.pipeline.actions.reaction_props.CgrCreated(in_column, out_column='CGRCreated')

Bases: object

Action for determining if a CGR can be created from the reaction smiles

Parameters:

• in_column (str)

• out_column (str)

in_column: str

pretty_name: ClassVar[str] = 'cgr_created'

out_column: str = 'CGRCreated'

class rxnutils.pipeline.actions.reaction_props.CgrNumberOfDynamicBonds(in_column, out_column='NDynamicBonds')

Bases: object

Action for calculating the number of dynamic bonds

Parameters:

• in_column (str)

• out_column (str)

in_column: str

pretty_name: ClassVar[str] = 'cgr_dynamic_bonds'

out_column: str = 'NDynamicBonds'

class rxnutils.pipeline.actions.reaction_props.ProductAtomMappingStats(in_column, unmapped_column='UnmappedProdAtoms', widow_column='WidowAtoms')

Bases: object

Action for collecting statistics of product atom mapping

Parameters:

• in_column (str)

• unmapped_column (str)

• widow_column (str)

pretty_name: ClassVar[str] = 'product_atommapping_stats'

in_column: str

unmapped_column: str = 'UnmappedProdAtoms'

widow_column: str = 'WidowAtoms'

class rxnutils.pipeline.actions.reaction_props.ProductSize(in_column, out_column='ProductSize')

Bases: object

Action for counting product size

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'productsize'

in_column: str

out_column: str = 'ProductSize'

class rxnutils.pipeline.actions.reaction_props.PseudoReactionHash(in_column, out_column='PseudoHash', no_reagents=False)

Bases: object

Action for creating a reaction hash based on InChI keys

Parameters:

• in_column (str)

• out_column (str)

• no_reagents (bool)

pretty_name: ClassVar[str] = 'pseudo_reaction_hash'

in_column: str

out_column: str = 'PseudoHash'

no_reagents: bool = False

class rxnutils.pipeline.actions.reaction_props.PseudoSmilesHash(in_column, out_column='PseudoSmilesHash')

Bases: object

Action for creating a reaction hash based on SMILES

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'pseudo_smiles_hash'

in_column: str

out_column: str = 'PseudoSmilesHash'

class rxnutils.pipeline.actions.reaction_props.ReactantProductAtomBalance(in_column, out_column='RxnAtomBalance')

Bases: object

Action for computing atom balance

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'atombalance'

in_column: str

out_column: str = 'RxnAtomBalance'

class rxnutils.pipeline.actions.reaction_props.ReactantSize(in_column, out_column='ReactantSize')

Bases: object

Action for counting reactant size

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'reactantsize'

in_column: str

out_column: str = 'ReactantSize'

class rxnutils.pipeline.actions.reaction_props.MaxRingNumber(in_column, out_column='MaxRings')

Bases: object

Action for calculating the maximum number of rings in either the product or reactant For a reaction without reactants or products, it will return 0 to enable easy arithmetic comparison

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'maxrings'

in_column: str

out_column: str = 'MaxRings'

class rxnutils.pipeline.actions.reaction_props.RingNumberChange(in_column, out_column='NRingChange')

Bases: object

Action for calculating if reaction has change in number of rings

A positive number from this action implies that a ring was formed during the reaction

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'ringnumberchange'

in_column: str

out_column: str = 'NRingChange'

class rxnutils.pipeline.actions.reaction_props.RingBondMade(in_column, out_column='RingBondMade')

Bases: object

Action for flagging if reaction has made a ring bond in the product

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'ringbondmade'

in_column: str

out_column: str = 'RingBondMade'

class rxnutils.pipeline.actions.reaction_props.RingMadeSize(in_column, out_column='RingMadeSize')

Bases: object

Action for computing the size of a newly formed ring

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'ringmadesize'

in_column: str

out_column: str = 'RingMadeSize'

class rxnutils.pipeline.actions.reaction_props.SmilesLength(in_column, out_column='SmilesLength')

Bases: object

Action for counting SMILES length

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'smiles_length'

in_column: str

out_column: str = 'SmilesLength'

class rxnutils.pipeline.actions.reaction_props.SmilesSanitizable(in_column, out_column='SmilesSanitizable')

Bases: object

Action for checking if SMILES are sanitizable

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'smiles_sanitizable'

in_column: str

out_column: str = 'SmilesSanitizable'

class rxnutils.pipeline.actions.reaction_props.StereoInvention(in_column, out_column='StereoInvention')

Bases: object

Flags reactions where non-stereo compounds (No “@”s in SMILES) turn into stereo compounds (containing “@”)

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'stereo_invention'

in_column: str

out_column: str = 'StereoInvention'

class rxnutils.pipeline.actions.reaction_props.StereoCentreChanges(in_column, out_column='HasStereoChanges', stereo_changes_column='StereoChanges')

Bases: object

Action for checking if stereogenic centre in reaction center is changing during the reaction

Will create two columns:

1. A boolean column indicating True or False if it has stereochanges

2. A description of the stereo information before and after the reaction

Parameters:

• in_column (str)

• out_column (str)

• stereo_changes_column (str)

pretty_name: ClassVar[str] = 'stereo_centre_changes'

in_column: str

out_column: str = 'HasStereoChanges'

stereo_changes_column: str = 'StereoChanges'

class rxnutils.pipeline.actions.reaction_props.StereoHasChiralReagent(in_column, out_column='HasChiralReagent')

Bases: object

Action for checking if reagent has stereo centres

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'stereo_chiral_reagent'

in_column: str

out_column: str = 'HasChiralReagent'

class rxnutils.pipeline.actions.reaction_props.StereoCenterIsCreated(in_column='StereoChanges', out_column='StereoCentreCreated')

Bases: object

Action for checking if stereo centre is created during reaction

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'stereo_centre_created'

in_column: str = 'StereoChanges'

out_column: str = 'StereoCentreCreated'

class rxnutils.pipeline.actions.reaction_props.StereoCenterIsRemoved(in_column='StereoChanges', out_column='StereoCentreRemoved')

Bases: object

Action for checking if stereo centre is removed during reaction

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'stereo_centre_removed'

in_column: str = 'StereoChanges'

out_column: str = 'StereoCentreRemoved'

class rxnutils.pipeline.actions.reaction_props.StereoCenterInReactantPotential(in_column, out_column='PotentialStereoCentre')

Bases: object

Action for checking if there is a potential stereo centre in the reaction

Do not consider changes to bond stereochemistry

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'potential_stereo_center'

in_column: str

out_column: str = 'PotentialStereoCentre'

class rxnutils.pipeline.actions.reaction_props.StereoCenterOutsideReaction(in_column, out_column='StereoOutside')

Bases: object

Action for checking if there is a stereo centre outside the reaction centre

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'stereo_centre_outside'

in_column: str

out_column: str = 'StereoOutside'

class rxnutils.pipeline.actions.reaction_props.StereoMesoProduct(in_column, out_column='MesoProduct')

Bases: object

Action for checking if the product is a meso compound

Parameters:

• in_column (str)

• out_column (str)

pretty_name: ClassVar[str] = 'meso_product'

in_column: str

out_column: str = 'MesoProduct'

rxnutils.pipeline.actions.templates module

Module containing template validation actions

class rxnutils.pipeline.actions.templates.CountTemplateComponents(in_column, nreactants_column='nreactants', nreagents_column='nreagents', nproducts_column='nproducts')

Bases: object

Action for counting template components

Parameters:

• in_column (str)

• nreactants_column (str)

• nreagents_column (str)

• nproducts_column (str)

pretty_name: ClassVar[str] = 'count_template_components'

in_column: str

nreactants_column: str = 'nreactants'

nreagents_column: str = 'nreagents'

nproducts_column: str = 'nproducts'

class rxnutils.pipeline.actions.templates.RetroTemplateReproduction(template_column, smiles_column, expected_reactants_column='TemplateGivesTrueReactants', other_reactants_column='TemplateGivesOtherReactants', noutcomes_column='TemplateGivesNOutcomes')

Bases: object

Action for checking template reproduction

Parameters:

• template_column (str)

• smiles_column (str)

• expected_reactants_column (str)

• other_reactants_column (str)

• noutcomes_column (str)

pretty_name: ClassVar[str] = 'retro_template_reproduction'

template_column: str

smiles_column: str

expected_reactants_column: str = 'TemplateGivesTrueReactants'

other_reactants_column: str = 'TemplateGivesOtherReactants'

noutcomes_column: str = 'TemplateGivesNOutcomes'

Module contents