forked from hpcaitech/FastFold
-
Notifications
You must be signed in to change notification settings - Fork 0
/
inference.sh
executable file
·20 lines (19 loc) · 956 Bytes
/
inference.sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
# add '--gpus [N]' to use N gpus for inference
# add '--enable_workflow' to use parallel workflow for data processing
# add '--use_precomputed_alignments [path_to_alignments]' to use precomputed msa
# add '--chunk_size [N]' to use chunk to reduce peak memory
# add '--inplace' to use inplace to save memory
python inference.py target.fasta data/pdb_mmcif/mmcif_files \
--output_dir ./outputs \
--gpus 2 \
--uniref90_database_path data/uniref90/uniref90.fasta \
--mgnify_database_path data/mgnify/mgy_clusters_2022_05.fa \
--pdb70_database_path data/pdb70/pdb70 \
--uniref30_database_path data/uniref30/UniRef30_2021_03 \
--bfd_database_path data/bfd/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt \
--jackhmmer_binary_path `which jackhmmer` \
--hhblits_binary_path `which hhblits` \
--hhsearch_binary_path `which hhsearch` \
--kalign_binary_path `which kalign` \
--enable_workflow \
--inplace