Skip to content

Commit

Permalink
Merge pull request #146 from JoseEspinosa/updates
Browse files Browse the repository at this point in the history 
Fix running only first sample when db downloaded
  • Loading branch information
@JoseEspinosa
JoseEspinosa authored Apr 18, 2024
2 parents 13b873e + 067d572 commit df3400f
Show file tree
Hide file tree
Showing 9 changed files with 103 additions and 102 deletions.
1 change: 1 addition & 0 deletions CHANGELOG.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -29,6 +29,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
- [[#135](https://github.com/nf-core/proteinfold/issues/135)] - Reduce Alphafold Docker images sizes.
- [[#115](https://github.com/nf-core/proteinfold/issues/115)] - Throw message error when profile conda is used.
- [[#131](https://github.com/nf-core/proteinfold/issues/131)] - Add esmfold small tests.
- [[#144](https://github.com/nf-core/proteinfold/issues/144)] - Force value channels when providing dbs (downloaded) in `main.nf` to enable the processing of multiple samples.

## 1.0.0 - White Silver Reebok

Expand Down
24 changes: 12 additions & 12 deletions conf/dbs.config
View file
Original file line number Diff line number Diff line change
Expand Up @@ -9,18 +9,18 @@
params {

// AlphaFold2 links
bfd = 'https://storage.googleapis.com/alphafold-databases/casp14_versions/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt.tar.gz'
small_bfd = 'https://storage.googleapis.com/alphafold-databases/reduced_dbs/bfd-first_non_consensus_sequences.fasta.gz'
alphafold2_params = 'https://storage.googleapis.com/alphafold/alphafold_params_2022-12-06.tar'
mgnify = 'https://storage.googleapis.com/alphafold-databases/casp14_versions/mgy_clusters_2018_12.fa.gz'
pdb70 = 'http://wwwuser.gwdg.de/~compbiol/data/hhsuite/databases/hhsuite_dbs/old-releases/pdb70_from_mmcif_200916.tar.gz'
pdb_mmcif = 'rsync.rcsb.org::ftp_data/structures/divided/mmCIF/' //'rsync.rcsb.org::ftp_data/structures/divided/mmCIF/' ftp.pdbj.org::ftp_data/structures/divided/mmCIF/ rsync.ebi.ac.uk::pub/databases/pdb/data/structures/divided/mmCIF/
pdb_obsolete = 'ftp://ftp.wwpdb.org/pub/pdb/data/status/obsolete.dat'
uniref30_alphafold2 = 'https://storage.googleapis.com/alphafold-databases/v2.3/UniRef30_2021_03.tar.gz'
uniref90 = 'ftp://ftp.uniprot.org/pub/databases/uniprot/uniref/uniref90/uniref90.fasta.gz'
pdb_seqres = 'ftp://ftp.wwpdb.org/pub/pdb/derived_data/pdb_seqres.txt'
uniprot_sprot = 'ftp://ftp.ebi.ac.uk/pub/databases/uniprot/current_release/knowledgebase/complete/uniprot_sprot.fasta.gz'
uniprot_trembl = 'ftp://ftp.ebi.ac.uk/pub/databases/uniprot/current_release/knowledgebase/complete/uniprot_trembl.fasta.gz'
bfd_link = 'https://storage.googleapis.com/alphafold-databases/casp14_versions/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt.tar.gz'
small_bfd_link = 'https://storage.googleapis.com/alphafold-databases/reduced_dbs/bfd-first_non_consensus_sequences.fasta.gz'
alphafold2_params_link = 'https://storage.googleapis.com/alphafold/alphafold_params_2022-12-06.tar'
mgnify_link = 'https://storage.googleapis.com/alphafold-databases/casp14_versions/mgy_clusters_2018_12.fa.gz'
pdb70_link = 'http://wwwuser.gwdg.de/~compbiol/data/hhsuite/databases/hhsuite_dbs/old-releases/pdb70_from_mmcif_200916.tar.gz'
pdb_mmcif_link = 'rsync.rcsb.org::ftp_data/structures/divided/mmCIF/' //'rsync.rcsb.org::ftp_data/structures/divided/mmCIF/' ftp.pdbj.org::ftp_data/structures/divided/mmCIF/ rsync.ebi.ac.uk::pub/databases/pdb/data/structures/divided/mmCIF/
pdb_obsolete_link = 'ftp://ftp.wwpdb.org/pub/pdb/data/status/obsolete.dat'
uniref30_alphafold2_link = 'https://storage.googleapis.com/alphafold-databases/v2.3/UniRef30_2021_03.tar.gz'
uniref90_link = 'ftp://ftp.uniprot.org/pub/databases/uniprot/uniref/uniref90/uniref90.fasta.gz'
pdb_seqres_link = 'ftp://ftp.wwpdb.org/pub/pdb/derived_data/pdb_seqres.txt'
uniprot_sprot_link = 'ftp://ftp.ebi.ac.uk/pub/databases/uniprot/current_release/knowledgebase/complete/uniprot_sprot.fasta.gz'
uniprot_trembl_link = 'ftp://ftp.ebi.ac.uk/pub/databases/uniprot/current_release/knowledgebase/complete/uniprot_trembl.fasta.gz'

// Alphafold paths
bfd_path = "${params.alphafold2_db}/bfd/*"
Expand Down
44 changes: 27 additions & 17 deletions main.nf
View file
Original file line number Diff line number Diff line change
Expand Up @@ -39,7 +39,7 @@ include { getColabfoldAlphafold2ParamsPath } from './subworkflows/local/utils_nf
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
*/

params.colabfold_alphafold2_params = getColabfoldAlphafold2Params()
params.colabfold_alphafold2_params_link = getColabfoldAlphafold2Params()
params.colabfold_alphafold2_params_path = getColabfoldAlphafold2ParamsPath()

/*
Expand Down Expand Up @@ -77,8 +77,18 @@ workflow NFCORE_PROTEINFOLD {
params.uniref90_path,
params.pdb_seqres_path,
params.uniprot_path,
params.uniprot_sprot,
params.uniprot_trembl
params.bfd_link,
params.small_bfd_link,
params.alphafold2_params_link,
params.mgnify_link,
params.pdb70_link,
params.pdb_mmcif_link,
params.pdb_obsolete_link,
params.uniref30_alphafold2_link,
params.uniref90_link,
params.pdb_seqres_link,
params.uniprot_sprot_link,
params.uniprot_trembl_link
)
ch_versions = ch_versions.mix(PREPARE_ALPHAFOLD2_DBS.out.versions)

Expand All @@ -90,16 +100,16 @@ workflow NFCORE_PROTEINFOLD {
params.full_dbs,
params.alphafold2_mode,
params.alphafold2_model_preset,
PREPARE_ALPHAFOLD2_DBS.out.params,
PREPARE_ALPHAFOLD2_DBS.out.bfd.ifEmpty([]),
PREPARE_ALPHAFOLD2_DBS.out.small_bfd.ifEmpty([]),
PREPARE_ALPHAFOLD2_DBS.out.mgnify,
PREPARE_ALPHAFOLD2_DBS.out.pdb70,
PREPARE_ALPHAFOLD2_DBS.out.pdb_mmcif,
PREPARE_ALPHAFOLD2_DBS.out.uniref30,
PREPARE_ALPHAFOLD2_DBS.out.uniref90,
PREPARE_ALPHAFOLD2_DBS.out.pdb_seqres,
PREPARE_ALPHAFOLD2_DBS.out.uniprot
PREPARE_ALPHAFOLD2_DBS.out.params.first(),
PREPARE_ALPHAFOLD2_DBS.out.bfd.ifEmpty([]).first(),
PREPARE_ALPHAFOLD2_DBS.out.small_bfd.ifEmpty([]).first(),
PREPARE_ALPHAFOLD2_DBS.out.mgnify.first(),
PREPARE_ALPHAFOLD2_DBS.out.pdb70.first(),
PREPARE_ALPHAFOLD2_DBS.out.pdb_mmcif.first(),
PREPARE_ALPHAFOLD2_DBS.out.uniref30.first(),
PREPARE_ALPHAFOLD2_DBS.out.uniref90.first(),
PREPARE_ALPHAFOLD2_DBS.out.pdb_seqres.first(),
PREPARE_ALPHAFOLD2_DBS.out.uniprot.first()
)
ch_multiqc = ALPHAFOLD2.out.multiqc_report
ch_versions = ch_versions.mix(ALPHAFOLD2.out.versions)
Expand All @@ -118,7 +128,7 @@ workflow NFCORE_PROTEINFOLD {
params.colabfold_alphafold2_params_path,
params.colabfold_db_path,
params.uniref30_colabfold_path,
params.colabfold_alphafold2_params,
params.colabfold_alphafold2_params_link,
params.colabfold_db_link,
params.uniref30_colabfold_link,
params.create_colabfold_index
Expand All @@ -131,9 +141,9 @@ workflow NFCORE_PROTEINFOLD {
COLABFOLD (
ch_versions,
params.colabfold_model_preset,
PREPARE_COLABFOLD_DBS.out.params,
PREPARE_COLABFOLD_DBS.out.colabfold_db,
PREPARE_COLABFOLD_DBS.out.uniref30,
PREPARE_COLABFOLD_DBS.out.params.first(),
PREPARE_COLABFOLD_DBS.out.colabfold_db.first(),
PREPARE_COLABFOLD_DBS.out.uniref30.first(),
params.num_recycle
)
ch_multiqc = COLABFOLD.out.multiqc_report
Expand Down
19 changes: 0 additions & 19 deletions modules.json
View file
Original file line number Diff line number Diff line change
Expand Up @@ -38,25 +38,6 @@
}
}
},
"subworkflows": {
"nf-core": {
"utils_nextflow_pipeline": {
"branch": "master",
"git_sha": "cd08c91373cd00a73255081340e4914485846ba1",
"installed_by": ["subworkflows"]
},
"utils_nfcore_pipeline": {
"branch": "master",
"git_sha": "262b17ed2aad591039f914951659177e6c39a8d8",
"installed_by": ["subworkflows"]
},
"utils_nfvalidation_plugin": {
"branch": "master",
"git_sha": "cd08c91373cd00a73255081340e4914485846ba1",
"installed_by": ["subworkflows"]
}
}
},
"subworkflows": {
"nf-core": {
"utils_nextflow_pipeline": {
Expand Down
24 changes: 12 additions & 12 deletions nextflow.config
View file
Original file line number Diff line number Diff line change
Expand Up @@ -22,18 +22,18 @@ params {
alphafold2_db = null

// Alphafold2 links
bfd = null
small_bfd = null
alphafold2_params = null
mgnify = null
pdb70 = null
pdb_mmcif = null
pdb_obsolete = null
uniref30_alphafold2 = null
uniref90 = null
pdb_seqres = null
uniprot_sprot = null
uniprot_trembl = null
bfd_link = null
small_bfd_link = null
alphafold2_params_link = null
mgnify_link = null
pdb70_link = null
pdb_mmcif_link = null
pdb_obsolete_link = null
uniref30_alphafold2_link = null
uniref90_link = null
pdb_seqres_link = null
uniprot_sprot_link = null
uniprot_trembl_link = null

// Alphafold2 paths
bfd_path = null
Expand Down
26 changes: 13 additions & 13 deletions nextflow_schema.json
View file
Original file line number Diff line number Diff line change
Expand Up @@ -283,73 +283,73 @@
"fa_icon": "fas fa-database",
"description": "Parameters used to provide the links to the DBs and parameters public resources to Alphafold2.",
"properties": {
"bfd": {
"bfd_link": {
"type": "string",
"default": "https://storage.googleapis.com/alphafold-databases/casp14_versions/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt.tar.gz",
"description": "Link to BFD dababase",
"fa_icon": "fas fa-link"
},
"small_bfd": {
"small_bfd_link": {
"type": "string",
"default": "https://storage.googleapis.com/alphafold-databases/reduced_dbs/bfd-first_non_consensus_sequences.fasta.gz",
"description": "Link to a reduced version of the BFD dababase",
"fa_icon": "fas fa-link"
},
"alphafold2_params": {
"alphafold2_params_link": {
"type": "string",
"default": "https://storage.googleapis.com/alphafold/alphafold_params_2022-12-06.tar",
"description": "Link to the Alphafold2 parameters",
"fa_icon": "fas fa-link"
},
"mgnify": {
"mgnify_link": {
"type": "string",
"default": "https://storage.googleapis.com/alphafold-databases/casp14_versions/mgy_clusters_2018_12.fa.gz",
"description": "Link to the MGnify database",
"fa_icon": "fas fa-link"
},
"pdb70": {
"pdb70_link": {
"type": "string",
"default": "http://wwwuser.gwdg.de/~compbiol/data/hhsuite/databases/hhsuite_dbs/old-releases/pdb70_from_mmcif_200916.tar.gz",
"description": "Link to the PDB70 database",
"fa_icon": "fas fa-link"
},
"pdb_mmcif": {
"pdb_mmcif_link": {
"type": "string",
"default": "rsync.rcsb.org::ftp_data/structures/divided/mmCIF/",
"description": "Link to the PDB mmCIF database",
"fa_icon": "fas fa-link"
},
"pdb_obsolete": {
"pdb_obsolete_link": {
"type": "string",
"default": "ftp://ftp.wwpdb.org/pub/pdb/data/status/obsolete.dat",
"description": "Link to the PDV obsolete database",
"fa_icon": "fas fa-link"
},
"uniref30_alphafold2": {
"uniref30_alphafold2_link": {
"type": "string",
"default": "https://storage.googleapis.com/alphafold-databases/v2.3/UniRef30_2021_03.tar.gz",
"description": "Link to the Uniclust30 database",
"fa_icon": "fas fa-link"
},
"uniref90": {
"uniref90_link": {
"type": "string",
"default": "ftp://ftp.uniprot.org/pub/databases/uniprot/uniref/uniref90/uniref90.fasta.gz",
"description": "Link to the UniRef90 database",
"fa_icon": "fas fa-link"
},
"pdb_seqres": {
"pdb_seqres_link": {
"type": "string",
"default": "ftp://ftp.wwpdb.org/pub/pdb/derived_data/pdb_seqres.txt",
"description": "Link to the PDB SEQRES database",
"fa_icon": "fas fa-link"
},
"uniprot_sprot": {
"uniprot_sprot_link": {
"type": "string",
"default": "ftp://ftp.ebi.ac.uk/pub/databases/uniprot/current_release/knowledgebase/complete/uniprot_sprot.fasta.gz",
"description": "Link to the SwissProt UniProt database",
"fa_icon": "fas fa-link"
},
"uniprot_trembl": {
"uniprot_trembl_link": {
"type": "string",
"default": "ftp://ftp.ebi.ac.uk/pub/databases/uniprot/current_release/knowledgebase/complete/uniprot_trembl.fasta.gz",
"description": "Link to the TrEMBL UniProt database",
Expand Down Expand Up @@ -433,7 +433,7 @@
"description": "Link to the UniRef30 database",
"fa_icon": "fas fa-link"
},
"colabfold_alphafold2_params": {
"colabfold_alphafold2_params_link": {
"type": "string",
"description": "Link to the Alphafold2 parameters for Colabfold",
"fa_icon": "fas fa-link"
Expand Down
Loading

0 comments on commit df3400f

Please sign in to comment.