o ॵi8 @sdZddlZddlZddlmZedejZejddGdddZ d e d e fd d Z d e d ee fddZ d e d e fddZ dS)zDUtilities for extracting identifiers from MSA sequence descriptions.N)Optionala ^ # UniProtKB/TrEMBL or UniProtKB/Swiss-Prot (?:tr|sp) \| # A primary accession number of the UniProtKB entry. (?P[A-Za-z0-9]{6,10}) # Occasionally there is a _0 or _1 isoform suffix, which we ignore. (?:_\d)? \| # TREMBL repeats the accession ID here. Swiss-Prot has a mnemonic # protein ID code. (?:[A-Za-z0-9]+) _ # A mnemonic species identification code. (?P([A-Za-z0-9]){1,5}) # Small BFD uses a final value after an underscore, which we ignore. (?:_\d+)? $ T)frozenc@seZdZUdZeed<dS) Identifiers species_idN)__name__ __module__ __qualname__rstr__annotations__r r i/home/ubuntu/.local/lib/python3.10/site-packages/modelscope/models/science/unifold/msa/msa_identifiers.pyr0s rmsa_sequence_identifierreturncCs*tt|}|rt|ddStS)aGets accession id and species from an msa sequence identifier. The sequence identifier has the format specified by _UNIPROT_TREMBL_ENTRY_NAME_PATTERN or _UNIPROT_SWISSPROT_ENTRY_NAME_PATTERN. An example of a sequence identifier: `tr|A0A146SKV9|A0A146SKV9_FUNHE` Args: msa_sequence_identifier: a sequence identifier. Returns: An `Identifiers` instance with a species_id. These can be empty in the case where no identifier was found. SpeciesIdentifier)r)research_UNIPROT_PATTERNstriprgroup)rmatchesr r r _parse_sequence_identifier5sr descriptioncCs"|}|r|dddSdS)zHExtracts sequence identifier from description. Returns None if no match.r/N)split partition)rsplit_descriptionr r r _extract_sequence_identifierIsrcCst|}|dur tSt|S)z1Computes extra MSA features from the description.N)rrr)rsequence_identifierr r r get_identifiersRsr)__doc__ dataclassesrtypingrcompileVERBOSEr dataclassrr rrrr r r r s