Skip to content

Name manipulation

This package contains utilities that automatically attempt to transform or correct input names if the initial name-to-SMILES conversion fails. At the moment there are three major capabilities:

  • Chemical name correction (primarily aimed at OCR/typo artifacts, with validation via OPSIN)
  • Splitting on delimiters (e.g. 'BH3•THF' → 'BH3' and 'THF')
  • Peptide shorthand expansion (e.g. Boc-Ala-Gly-OMe → tert-butoxycarbonyl-l-alanyl-glycine methyl ester)

Each of these name manipulations is enabled by default, but can be disabled by setting the corresponding argument to False.

Chemical name correction

Applied when a chemical name looks like it might have OCR/typo artifacts, with validation via OPSIN.

  • Examples of artifacts:
  • 3,4-dihydro- 3-hydroxy-4-oxo-l,2,3-benzotriazine → 3,4-dihydro- 3-hydroxy-4-oxo-1,2,3-benzotriazine
  • ethyl perfiuorobutyrate → ethyl perfluorobutyrate
  • l-mercapto-Z-thiapropane → 1-mercapto-2-thiapropane
from cholla_chem import resolve_compounds_to_smiles

resolved_smiles = resolve_compounds_to_smiles(compounds_list=['l-mercapto-Z-thiapropane'], attempt_name_correction=True)

"{'l-mercapto-Z-thiapropane': 'SCSC'}"

Splitting on delimiters

Split chemical names on common delimiters and then solve each component separately.

from cholla_chem import resolve_compounds_to_smiles

resolved_smiles = resolve_compounds_to_smiles(compounds_list=['BH3•THF'], split_names_to_solve=True)

"{'BH3•THF': 'B.C1CCOC1'}"

Peptide shorthand expansion

Applied when a chemical name looks like peptide shorthand, typically containing amino-acid abbreviations delimited by hyphens, such as:

  • Ala-Gly-Ser
  • Boc-Asp-Lys(Boc)-OMe
  • cyclo(Ala-Gly-Ser) or cyclo[Ala-Gly-Ser]
from cholla_chem import resolve_compounds_to_smiles

resolved_smiles = resolve_compounds_to_smiles(compounds_list=['cyclo(Asp-Arg-Val-Tyr-Ile-His-Pro-Phe)'], resolve_peptide_shorthand=True)

"{'cyclo(Asp-Arg-Val-Tyr-Ile-His-Pro-Phe)': 'BN1[C@@H](CC(=O)O)C(=O)N[C@@H](CCCNC(N)=N)C(=O)N[C@@H](C(C)C)C(=O)N[C@@H](CC2=CC=C(C=C2)O)C(=O)N[C@@H]([C@@H](C)CC)C(=O)N[C@@H](CC2=CNC=N2)C(=O)N2[C@@H](CCC2)C(=O)N[C@@H](CC2=CC=CC=C2)C1=O'}"