Skip to content

README.workflow.md

Latest commit

 

History

History
310 lines (178 loc) · 6.22 KB

README.workflow.md

File metadata and controls

310 lines (178 loc) · 6.22 KB

This is the initial overview for the CATH AlphaFold NextFlow workflow.

1a. Global dataset (AF2 sequences)

Name: CREATE_ALL_AF2_CHAIN_FASTA

Input:

  • PARAM:AF2_DOWNLOAD_URL -- AlphaFold Download (gs://public-datasets-deepmind-alphafold/sequences.fasta)

Output:

  • FILE:ALL_AF2_CHAIN_FASTA

1b. Global dataset (Foldseek S95 library)

Name: CREATE_DOMAIN_S95_PDB_DIR

Input:

  • FILE:DOMAIN_LIST_S95
  • DIRECTORY:CATH_PDB_DOMAINS -- v4.3
  • PARAM:S95_DIR -- folder containing S95 PDB files

Output:

  • DIRECTORY:FOLDSEEK_S95_PDB_DB

Process:

  • rsync --files-from <FILE:DOMAIN_LIST_S95> <DIRECTORY:CATH_PDB_DOMAINS> <DIRECTORY:FOLDSEEK_S95_PDB_DB>

1c. Global dataset (Foldseek S95 library)

Name: CREATE_FOLDSEEK_S95_LIBRARY

Input:

  • PARAM:S95_DIR -- folder containing S95 PDB files

Output:

  • FILE:FOLDSEEK_S95_LIBRARY

Process:

  • foldseek createdb <s95_dir> <s95_db>

Notes:

  • currently in /SAN/cath/cath_v4_3_0/databases/foldseek/cath_s95/

2a. Local dataset (UniProt IDs)

Name: CREATE_DATASET_UNIPROT_IDS

Input:

  • SOURCE:OracleDB -- GENE3D_21.CATH_DOMAIN_PREDICTIONS{_EXTRA}
  • FILTER -- CONDITIONAL_EVALUE <= 1e-50, top 10k hits (ordered by increasing evalue)

Output:

  • FILE:CSV_UNIPROT_IDS -- UniProt_ID

2b. Local dataset (CATH domains predictions)

Name: CREATE_DATASET_CATH_FILES

Input:

  • SOURCE:OracleDB -- GENE3D_21.CATH_DOMAIN_PREDICTIONS{_EXTRA}
  • FILE:CSV_UNIPROT_IDS -- (UniProt_ID)

Output:

  • FILE:CSV_UNIPROT_MD5 -- (UniProt_ID, md5)
  • FILE:CRH_OUTPUT -- (md5, CATH_domain_ID, bitscore, boundaries, resolved_boundaries)
  • FILE:AF2_DOMAIN_LIST -- (AF_domain_ID_orig)
  • FILE:AF2_CHAIN_LIST -- (AF_chain_ID)
  • FILE:AF2_CATH_ORIG_ANNOTATIONS -- (AF_domain_ID_orig, CATH_domain_ID, UniProt_ID, md5, bitscore, resolved_boundaries, sfam_id, class_id)

2c. Local dataset (AF2 Structure)

Name: CREATE_DATASET_AF2_FILES

Input:

  • FILE:CSV_UNIPROT_MD5
  • SOURCE:HPC_ENVIRONMENT -- Google Cloud Storage (GCS), Computer Science (CS)
  • FILE:AF2_CHAIN_LIST

Output:

  • FILE:ALL_CHAIN_FASTA -- one FASTA file containing all AF2 chains
  • DIRECTORY:AF2_CHAIN_MMCIF -- one mmCIF file per AF2 chain

Process:

  • GCS -- cat [manifest file] | gsutil -m cp -I .
  • CS -- (local): symlinking (check if it exists)

3a. Run SETH (Chain sequence)

Name: CREATE_ANNOTATION_CHAIN_DISORDER

Input:

  • FILE:ALL_CHAIN_FASTA

Output:

  • FILE:SETH_CHAIN_OUTPUT_FILE -- Per-residue SETH scores (disordered 0 -> 1 ordered)

Process:

  • SETH_1.py -i <your input fasta file name> -o <the desired name of your output file>

3b. Filter (Disordered Chain)

Name: FILTER_DOMAINLIST_BY_CHAIN_DISORDER

Input:

  • FILE:SETH_CHAIN_OUTPUT_FILE
  • FILTER -- (SETH_GLOBAL_CHAIN_DISORDER, SETH_LOCAL_DOMAIN_DISORDER)
  • FILE:AF2_DOMAIN_LIST

Output:

  • FILE:AF2_DOMAIN_LIST
  • FILE:SETH_GLOBAL_CHAIN_DISORDER -- (AF_domain_ID_orig, seth_global_disorder_score)

Process:

  • python filter_disorder.py

4. Optimize domain boundaries (chop tails)

Name: OPTIMISE_DOMAIN_BOUNDARIES

Input:

  • FILE:AF2_CATH_ORIG_ANNOTATIONS
  • FILE:AF2_CHAIN_MMCIF
  • FILE:AF2_DOMAIN_LIST

Output:

  • FILE:AF2_DOMAIN_LIST_POST_TAILCHOP -- (AF_domain_ID_tailchop)
  • FILE:AF2_DOMAIN_MAPPING_POST_TAILCHOP -- (AF_domain_ID_orig, AF_domain_ID_tailchop)

Process:

  • python optimize_boundaries.py

5a. Run SETH (Domain sequence)

Name: CREATE_ANNOTATION_DOMAIN_DISORDER

Input:

  • FILE:ALL_CHAIN_FASTA
  • FILE:AF2_DOMAIN_LIST_POST_TAILCHOP

Output:

  • FILE:SETH_DOMAIN_OUTPUT_FILE -- Per-residue SETH scores (disordered 0 -> 1 ordered)

Process:

  • SETH_1.py -i <your input fasta file name> -o <the desired name of your output file>

5b. Filter (Disordered Domain)

Name: FILTER_DOMAINLIST_BY_DISORDER

Input:

  • FILE:SETH_DOMAIN_OUTPUT_FILE
  • FILTER -- (SETH_GLOBAL_CHAIN_DISORDER, SETH_LOCAL_DOMAIN_DISORDER)
  • FILE:AF2_DOMAIN_LIST_POST_TAILCHOP

Output:

  • FILE:AF2_DOMAIN_LIST_POST_DOMAIN_DISORDER
  • FILE:AF2_SETH_ANNOTATIONS -- (AF_domain_ID_tailchop, seth_global_disorder_score)

Process:

  • python filter_disorder.py

6. Filter: AF2 Quality (pLDDT, LUR)

Name: FILTER_DOMAINLIST_BY_AF2_QUALITY

Input:

  • FILE:AF2_CHAIN_MMCIF
  • FILE:AF2_DOMAIN_LIST_POST_DOMAIN_DISORDER

Output:

  • FILE:AF2_DOMAIN_LIST_POST_AF_QUALITY
  • FILE:AF2_QUALITY_ANNOTATIONS -- (AF_domain_ID_tailchop, PLDDT_average, LUR_score)

Process:

  • python filter_af2_quality.py

7. Filter: AF2 Packing (packing, Surf/Vol)

Name: FILTER_DOMAINLIST_BY_AF2_PACKING

Input:

  • FILE:AF2_CHAIN_MMCIF
  • FILE:AF2_DOMAIN_LIST_POST_AF_QUALITY

Output:

  • FILE:AF2_DOMAIN_LIST_POST_AF_PACKING
  • FILE:AF2_PACKING_ANNOTATIONS -- (AF_domain_ID_tailchop, packing_score, surf_vol_score)

Process:

  • python filter_af2_packing.py

8. Filter: SSE (DSSP+secmake)

Name: FILTER_DOMAINLIST_BY_SSE

Input:

  • FILE:AF2_CHAIN_MMCIF
  • FILE:AF2_DOMAIN_LIST_POST_AF_PACKING

Output:

  • FILE:AF2_DOMAIN_LIST_POST_SSE
  • FILE:AF2_SSE_ANNOTATIONS -- (AF_domain_ID_tailchop, sse_number)

Process:

  • python3 filter_sse.py

9. Run Chopping

Name: CHOP_AF2_DOMAINS

Input:

  • FILE:AF2_DOMAIN_LIST_POST_SSE -- af_<Uniprot_ID>/<start>-<stop>
  • DIRECTORY:AF2_CHAIN_MMCIF

Output:

  • DIRECTORY:AF2_DOMAIN_MMCIF
  • FILE:AF2_DOMAIN_MAPPING_POST_CHOPPING -- (AF_domain_ID_tailchop, AF_domain_ID_chopped)

Process:

submit.sh
/home/ucbcisi/work/2022_06_29.alphafold_rechop_corrected_domains/submit.sh

10. Foldseek S95

Name: CREATE_ANNOTATION_FOLDSEEK_S95

Input:

  • FILE:FOLDSEEK_S95_DB
  • FILE:AF2_DOMAIN_LIST_POST_SSE
  • DIRECTORY:AF2_DOMAIN_MMCIF

Output:

  • FILE:AF2_FOLDSEEK_ANNOTATIONS -- (AF_domain_ID_tailchop, foldseek_bitscore, foldseek_overlap)

Process:

foldseek createdb AF2_DOMAIN_MMCIF AF2_DOMAIN_MMCIF_DB
foldseek search AF2_DOMAIN_MMCIF_DB FOLDSEEK_S95_DB
foldseek convertalis AF2_DOMAIN_MMCIF_DB FOLDSEEK_S95_DB

11. Create Table

Name: CREATE_RESULTS_TABLE

Input:

  • FILE:AF2_DOMAIN_LIST_POST_SSE
  • FILE:AF2_SETH_ANNOTATIONS
  • FILE:AF2_QUALITY_ANNOTATIONS
  • FILE:AF2_PACKING_ANNOTATIONS
  • FILE:AF2_CATH_ORIG_ANNOTATIONS
  • FILE:AF2_SSE_ANNOTATIONS
  • FILE:AF2_FOLDSEEK_ANNOTATIONS

Output:

  • FILE:CATH_AF2_TABLE

Process:

  • python collate_data_to_table.py