Merge pull request #156 from JoseEspinosa/updates

Update DBs URLs  and Dockerfiles

CHANGELOG.md

@@ -34,6 +34,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - [[#145](https://github.com/nf-core/proteinfold/issues/145)] - Implement test to check the processes/subworkflows triggered when downloading the databases.
 - [[#130](https://github.com/nf-core/proteinfold/issues/130)] - Add `--skip_multiqc` parameter.
 - [[PR #154](https://github.com/nf-core/proteinfold/pull/154)] - Update pipeline template to [nf-core/tools 2.14.1](https://github.com/nf-core/tools/releases/tag/2.14.1).
+- [[#148](https://github.com/nf-core/proteinfold/issues/148)] - Update Colabfold DBs.
 
 ### Parameters
 

conf/dbs.config

@@ -11,16 +11,16 @@ params {
     // AlphaFold2 links
     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'
+    alphafold2_params_link   = 'https://storage.googleapis.com/alphafold/alphafold_params_2022-12-06.tar'
+    mgnify_link              = 'https://storage.googleapis.com/alphafold-databases/v2.3/mgy_clusters_2022_05.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'
+    pdb_mmcif_link           = 'rsync.rcsb.org::ftp_data/structures/divided/mmCIF/' //Other sources available: '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        = 'https://files.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'
+    uniref90_link            = 'https://ftp.ebi.ac.uk/pub/databases/uniprot/uniref/uniref90/uniref90.fasta.gz'
+    pdb_seqres_link          = 'https://files.wwpdb.org/pub/pdb/derived_data/pdb_seqres.txt'
+    uniprot_sprot_link       = 'https://ftp.ebi.ac.uk/pub/databases/uniprot/current_release/knowledgebase/complete/uniprot_sprot.fasta.gz'
+    uniprot_trembl_link      = 'https://ftp.ebi.ac.uk/pub/databases/uniprot/current_release/knowledgebase/complete/uniprot_trembl.fasta.gz'
 
     // Alphafold paths
     bfd_path                 = "${params.alphafold2_db}/bfd/*"
@@ -36,11 +36,11 @@ params {
 
     // Colabfold links
     colabfold_db_link       = 'http://wwwuser.gwdg.de/~compbiol/colabfold/colabfold_envdb_202108.tar.gz'
-    uniref30_colabfold_link = 'https://wwwuser.gwdg.de/~compbiol/colabfold/uniref30_2202.tar.gz'
+    uniref30_colabfold_link = 'https://wwwuser.gwdg.de/~compbiol/colabfold/uniref30_2302.tar.gz'
 
     // Colabfold paths
     colabfold_db_path                = "${params.colabfold_db}/colabfold_envdb_202108"
-    uniref30_colabfold_path          = "${params.colabfold_db}/uniref30_2202"
+    uniref30_colabfold_path          = "${params.colabfold_db}/uniref30_2302"
     colabfold_alphafold2_params_tags = [
         "alphafold2_multimer_v1" : "alphafold_params_colab_2021-10-27",
         "alphafold2_multimer_v2" : "alphafold_params_colab_2022-03-02",

dockerfiles/Dockerfile_nfcore-proteinfold_alphafold2_msa

@@ -9,7 +9,7 @@ LABEL authors="Luisa Santus, Athanasios Baltzis, Jose Espinosa-Carrasco, Leila M
 SHELL ["/bin/bash", "-o", "pipefail", "-c"]
 
 # Add env variables
-ENV LD_LIBRARY_PATH="/conda/lib:/usr/local/cuda-11.1/lib64:$LD_LIBRARY_PATH"
+ENV LD_LIBRARY_PATH="/conda/lib:/usr/local/cuda-11.4/lib64:$LD_LIBRARY_PATH"
 ENV PATH="/conda/bin:$PATH"
 
 RUN apt-key adv --keyserver keyserver.ubuntu.com --recv-keys A4B469963BF863CC

dockerfiles/Dockerfile_nfcore-proteinfold_alphafold2_split

@@ -9,7 +9,7 @@ LABEL authors="Athanasios Baltzis, Jose Espinosa-Carrasco, Leila Mansouri" \
 SHELL ["/bin/bash", "-o", "pipefail", "-c"]
 
 # Add env variables
-ENV LD_LIBRARY_PATH="/conda/lib:/usr/local/cuda-11.1/lib64:$LD_LIBRARY_PATH"
+ENV LD_LIBRARY_PATH="/conda/lib:/usr/local/cuda-11.4/lib64:$LD_LIBRARY_PATH"
 ENV PATH="/conda/bin:$PATH"
 
 RUN apt-key adv --keyserver keyserver.ubuntu.com --recv-keys A4B469963BF863CC

dockerfiles/Dockerfile_nfcore-proteinfold_alphafold2_standard

@@ -9,7 +9,7 @@ LABEL authors="Athanasios Baltzis, Jose Espinosa-Carrasco, Leila Mansouri" \
 SHELL ["/bin/bash", "-o", "pipefail", "-c"]
 
 # Add env variables
-ENV LD_LIBRARY_PATH="/conda/lib:/usr/local/cuda-11.1/lib64:$LD_LIBRARY_PATH"
+ENV LD_LIBRARY_PATH="/conda/lib:/usr/local/cuda-11.4/lib64:$LD_LIBRARY_PATH"
 ENV PATH="/conda/bin:$PATH"
 
 RUN apt-key adv --keyserver keyserver.ubuntu.com --recv-keys A4B469963BF863CC

dockerfiles/Dockerfile_nfcore-proteinfold_colabfold

@@ -1,5 +1,6 @@
-FROM nvidia/cuda:11.4.2-cudnn8-runtime-ubuntu18.04
-LABEL authors="Athanasios Baltzis, Leila Mansouri" \
+FROM nvidia/cuda:11.4.3-cudnn8-runtime-ubuntu18.04
+
+LABEL authors="Athanasios Baltzis, Jose Espinosa-Carrasco, Leila Mansouri" \
     title="nfcore/proteinfold_colabfold" \
     Version="1.1.0" \
     description="Docker image containing all software requirements to run the COLABFOLD_BATCH module using the nf-core/proteinfold pipeline"
@@ -24,7 +25,7 @@ RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y \
     && rm -rf /var/lib/apt/lists/*
 
 RUN cd / \
-    && wget https://raw.githubusercontent.com/YoshitakaMo/localcolabfold/fb0598ebbd8227096a052d3684a549ddf33d06bb/install_colabbatch_linux.sh \
+    && wget https://raw.githubusercontent.com/YoshitakaMo/localcolabfold/82a3635/install_colabbatch_linux.sh \
     && sed -i "/colabfold.download/d" install_colabbatch_linux.sh \
     && sed -i "s|cudatoolkit==.*\sopenmm|cudatoolkit==11.1.1 openmm|g" install_colabbatch_linux.sh \
     && bash install_colabbatch_linux.sh
@@ -33,4 +34,4 @@ RUN /localcolabfold/colabfold-conda/bin/python3.10 -m pip install tensorflow-cpu
 
 #Silence download of the AlphaFold2 params
 RUN  sed -i "s|download_alphafold_params(|#download_alphafold_params(|g" /localcolabfold/colabfold-conda/lib/python3.10/site-packages/colabfold/batch.py
-
+RUN  sed -i "s|if args\.num_models|#if args\.num_models|g" /localcolabfold/colabfold-conda/lib/python3.10/site-packages/colabfold/batch.py

modules/local/colabfold_batch.nf

@@ -7,7 +7,7 @@ process COLABFOLD_BATCH {
         error("Local COLABFOLD_BATCH module does not support Conda. Please use Docker / Singularity / Podman instead.")
     }
 
-    container "nf-core/proteinfold_colabfold:1.1.0"
+    container "nf-core/proteinfold_colabfold:dev"
 
     input:
     tuple val(meta), path(fasta)

modules/local/mmseqs_colabfoldsearch.nf

@@ -7,7 +7,7 @@ process MMSEQS_COLABFOLDSEARCH {
         error("Local MMSEQS_COLABFOLDSEARCH module does not support Conda. Please use Docker / Singularity / Podman instead.")
     }
 
-    container "nf-core/proteinfold_colabfold:1.1.0"
+    container "nf-core/proteinfold_colabfold:dev"
 
     input:
     tuple val(meta), path(fasta)
@@ -16,7 +16,7 @@ process MMSEQS_COLABFOLDSEARCH {
     path uniref30
 
     output:
-    tuple val(meta), path("${meta.id}.a3m"), emit: a3m
+    tuple val(meta), path("**.a3m"), emit: a3m
     path "versions.yml", emit: versions
 
     when:
@@ -36,8 +36,6 @@ process MMSEQS_COLABFOLDSEARCH {
         ./db \\
         "result/"
 
-    cp result/0.a3m ${meta.id}.a3m
-
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":
         colabfold_search: $VERSION
@@ -47,7 +45,8 @@ process MMSEQS_COLABFOLDSEARCH {
     stub:
     def VERSION = '1.5.2' // WARN: Version information not provided by tool on CLI. Please update this string when bumping container versions.
     """
-    touch ${meta.id}.a3m
+    mkdir results
+    touch results/${meta.id}.a3m
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":

nextflow_schema.json

@@ -316,7 +316,7 @@
                 },
                 "mgnify_link": {
                     "type": "string",
-                    "default": "https://storage.googleapis.com/alphafold-databases/casp14_versions/mgy_clusters_2018_12.fa.gz",
+                    "default": "https://storage.googleapis.com/alphafold-databases/v2.3/mgy_clusters_2022_05.fa.gz",
                     "description": "Link to the MGnify database",
                     "fa_icon": "fas fa-link"
                 },
@@ -334,8 +334,8 @@
                 },
                 "pdb_obsolete_link": {
                     "type": "string",
-                    "default": "ftp://ftp.wwpdb.org/pub/pdb/data/status/obsolete.dat",
-                    "description": "Link to the PDV obsolete database",
+                    "default": "https://files.wwpdb.org/pub/pdb/data/status/obsolete.dat",
+                    "description": "Link to the PDB obsolete database",
                     "fa_icon": "fas fa-link"
                 },
                 "uniref30_alphafold2_link": {
@@ -346,25 +346,25 @@
                 },
                 "uniref90_link": {
                     "type": "string",
-                    "default": "ftp://ftp.uniprot.org/pub/databases/uniprot/uniref/uniref90/uniref90.fasta.gz",
+                    "default": "https://ftp.ebi.ac.uk/pub/databases/uniprot/uniref/uniref90/uniref90.fasta.gz",
                     "description": "Link to the UniRef90 database",
                     "fa_icon": "fas fa-link"
                 },
                 "pdb_seqres_link": {
                     "type": "string",
-                    "default": "ftp://ftp.wwpdb.org/pub/pdb/derived_data/pdb_seqres.txt",
+                    "default": "https://files.wwpdb.org/pub/pdb/derived_data/pdb_seqres.txt",
                     "description": "Link to the PDB SEQRES database",
                     "fa_icon": "fas fa-link"
                 },
                 "uniprot_sprot_link": {
                     "type": "string",
-                    "default": "ftp://ftp.ebi.ac.uk/pub/databases/uniprot/current_release/knowledgebase/complete/uniprot_sprot.fasta.gz",
+                    "default": "https://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_link": {
                     "type": "string",
-                    "default": "ftp://ftp.ebi.ac.uk/pub/databases/uniprot/current_release/knowledgebase/complete/uniprot_trembl.fasta.gz",
+                    "default": "https://ftp.ebi.ac.uk/pub/databases/uniprot/current_release/knowledgebase/complete/uniprot_trembl.fasta.gz",
                     "description": "Link to the TrEMBL UniProt database",
                     "fa_icon": "fas fa-link"
                 }
@@ -442,7 +442,7 @@
                 },
                 "uniref30_colabfold_link": {
                     "type": "string",
-                    "default": "https://wwwuser.gwdg.de/~compbiol/colabfold/uniref30_2202.tar.gz",
+                    "default": "https://wwwuser.gwdg.de/~compbiol/colabfold/uniref30_2302.tar.gz",
                     "description": "Link to the UniRef30 database",
                     "fa_icon": "fas fa-link"
                 },