o ॵiL @sxddlZddlmZmZddlZddlmZddlmZdee dee fddZ d e d e dee dee fd d Z dS) N)MappingSequence)logging)protein sequences descriptionscCsg}|D] }||vr||qddttj|D}g}t||D]\}}tj||}||d|||q"||fS)zx Makes a mapping from PDB-format chain ID to sequence and description, and parses the order of multi-chains cSsi|] \}}|g|dqS))rsequence).0chain_idseqr r _/home/ubuntu/.local/lib/python3.10/site-packages/modelscope/models/science/unifold/msa/utils.py sz$get_chain_id_map..r)appendzipr PDB_CHAIN_IDSindex)rr unique_seqsr chain_id_map chain_orderdesr r r r get_chain_id_map s   r fasta_nameoutput_dir_basec Cst|t|krtdt|dt|dt|tjkr(tdt|dt||\}}tj||}tj|sAt |tj|d}t |d}t j ||dd d Wd n1sawYtj|d } t | d}| d |Wd n1swYtdd |g} g} |D]G} |d| } tj|| d}d| }|| d}t |d}| d|d|Wd n1swY| | | |q| | fS)zz Divides the multi-chains fasta into several single fasta files and records multi-chains mapping information. z7sequences and descriptions must have equal length. Got z != .z=Cannot process more chains than the PDB format supports. Got z chains.zchain_id_map.jsonwT)indent sort_keysNz chains.txt z/Mapping multi-chains fasta with chain order: %sz_{}z.fastazchain_{}r> )len ValueErrorrPDB_MAX_CHAINSrospathjoinexistsmakedirsopenjsondumpwriterinfokeysformatr)rrrrrr output_dirchain_id_map_pathfchain_order_path temp_names temp_pathsr temp_name temp_pathrr r r r divide_multi_chains+sP            r9) r%typingrrr+abslr&modelscope.models.science.unifold.datarstrrr9r r r r s&