Merge pull request #62 from nf-core/hmmsearch-funfams

Download and hmmsearch FunFam

Author: vagkaratzas

CHANGELOG.md

@@ -9,6 +9,7 @@ Initial release of nf-core/proteinannotator, created with the [nf-core](https://
 
 ### `Added`
 
+- [#62](https://github.com/nf-core/proteinannotator/pull/62) - Added the option to download and use the latest FunFam HMM library (or use path to an existing one) for domain annotation. (by @vagkaratzas)
 - [#61](https://github.com/nf-core/proteinannotator/pull/61) - Added nf-core modules `ARIA2` and `HMMER_HMMSEARCH` to download latest Pfam HMM library (or use path to existing one) and match domains to input sequences. (by @vagkaratzas)
 - [#60](https://github.com/nf-core/proteinannotator/pull/60) - Added nf-core module `S4PRED_RUNMODEL` for secondary structure prediction (i.e., α-helix, a β-strand or a coil). (by @vagkaratzas)
 - [#59](https://github.com/nf-core/proteinannotator/pull/59) - Added nf-core qc and pre-processing subworkflow for amino acid sequences `FAA_SEQFU_SEQKIT`. (by @vagkaratzas)

conf/modules.config

@@ -74,15 +74,23 @@ process {
         ]
     }
 
-    withName: 'NFCORE_PROTEINANNOTATOR:PROTEINANNOTATOR:DOMAIN_ANNOTATION:ARIA2' {
+    withName: 'NFCORE_PROTEINANNOTATOR:PROTEINANNOTATOR:DOMAIN_ANNOTATION:ARIA2_PFAM' {
         publishDir = [
             path: { "${params.outdir}/downloaded_dbs/" },
             mode: params.publish_dir_mode,
             saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
         ]
     }
 
-    withName: 'NFCORE_PROTEINANNOTATOR:PROTEINANNOTATOR:DOMAIN_ANNOTATION:HMMER_HMMSEARCH' {
+    withName: 'NFCORE_PROTEINANNOTATOR:PROTEINANNOTATOR:DOMAIN_ANNOTATION:ARIA2_FUNFAM' {
+        publishDir = [
+            path: { "${params.outdir}/downloaded_dbs/" },
+            mode: params.publish_dir_mode,
+            saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+        ]
+    }
+
+    withName: 'NFCORE_PROTEINANNOTATOR:PROTEINANNOTATOR:DOMAIN_ANNOTATION:HMMSEARCH_PFAM' {
         ext.args   = { "-E ${params.hmmsearch_evalue_cutoff}" }
         publishDir = [
             path: { "${params.outdir}/domain_annotation/pfam/" },
@@ -92,6 +100,16 @@ process {
         ]
     }
 
+    withName: 'NFCORE_PROTEINANNOTATOR:PROTEINANNOTATOR:DOMAIN_ANNOTATION:HMMSEARCH_FUNFAM' {
+        ext.args   = { "-E ${params.hmmsearch_evalue_cutoff}" }
+        publishDir = [
+            path: { "${params.outdir}/domain_annotation/funfam/" },
+            mode: params.publish_dir_mode,
+            pattern: "*.domtbl.gz",
+            saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+        ]
+    }
+
     withName: 'NFCORE_PROTEINANNOTATOR:PROTEINANNOTATOR:S4PRED_RUNMODEL' {
         ext.prefix = { params.s4pred_outfmt }
         ext.args   = { "--outfmt ${params.s4pred_outfmt}" }

conf/test.config

@@ -25,7 +25,8 @@ params {
     // Input data
     input = params.pipelines_testdata_base_path + 'proteinfold/testdata/samplesheet/v1.2/samplesheet.csv'
     // Domain annotation
-    pfam_latest_link = params.pipelines_testdata_base_path + 'proteinannotator/testdata/pfam/Pfam-A_test.hmm.gz'
+    pfam_latest_link   = params.pipelines_testdata_base_path + 'proteinannotator/testdata/pfam/Pfam-A_test.hmm.gz'
+    funfam_latest_link = params.pipelines_testdata_base_path + 'proteinannotator/testdata/funfam/funfam-hmm3-v4_3_0_test.lib.gz'
     // Functional annotation
     skip_interproscan = true
 }

conf/test_full.config

@@ -17,5 +17,6 @@ params {
     // Input data for full size test
     input = params.pipelines_testdata_base_path + 'proteinannotator/samplesheet/snap25-isoforms.csv'
     // Domain annotation
-    pfam_latest_link = params.pipelines_testdata_base_path + 'proteinannotator/testdata/pfam/Pfam-A_test.hmm.gz'
+    pfam_latest_link   = params.pipelines_testdata_base_path + 'proteinannotator/testdata/pfam/Pfam-A_test.hmm.gz'
+    funfam_latest_link = params.pipelines_testdata_base_path + 'proteinannotator/testdata/funfam/funfam-hmm3-v4_3_0_test.lib.gz'
 }

docs/output.md

@@ -6,8 +6,6 @@ This document describes the output produced by the pipeline. Most of the plots a
 
 The directories listed below will be created in the results directory after the pipeline has finished. All paths are relative to the top-level results directory.
 
-<!-- TODO nf-core: Write this documentation describing your workflow's output -->
-
 ## Pipeline overview
 
 The pipeline is built using [Nextflow](https://www.nextflow.io/) and processes data using the following steps:
@@ -17,8 +15,8 @@ The pipeline is built using [Nextflow](https://www.nextflow.io/) and processes d
   - [SeqKit](#seqkit) for preprocessing input amino acid sequences (i.e., gap removal, convert to upper case, validate, filter by length, replace special characters such as `/`, and remove duplicate sequences)
 
 - [Domain annotation](#domain-annotation) Annotate proteins with domains from established repositories.
-  - [aria2](#aria2) - To optionally download the latest Pfam database through the pipeline.
-  - [hmmer](#hmmer) - To optionally match the input sequence to known Pfam domains through `hmmer/hmmsearch`
+  - [aria2](#aria2) - To optionally download the latest Pfam and/or FunFam databases through the pipeline.
+  - [hmmer](#hmmer) - To optionally match the input sequence to known Pfam and/or FunFam domains through `hmmer/hmmsearch`
 
 - [Functional annotation](#functional-annotation) Annotate proteins with functional domains
   - [InterProScan](#Interproscan) - Search the InterPro database for functional domains
@@ -73,9 +71,12 @@ The `seqkit` module is used for initial preprocessing (i.e., gap removal, conver
 
 - `downloaded_dbs/`
   - `Pfam-A*.hmm.gz`: (optional) The latest full, or a minimal test, Pfam-A HMM database that can be downloaded through the pipeline.
+  - `funfam-hmm3-v4_3_0*.lib.gz`: (optional) The latest (v4_3_0) full, or a minimal test, FunFam HMM database that can be downloaded through the pipeline.
 
 </details>
 
+If the `skip_*` flags (e.g., `skip_pfam`, `skip_funfam`) for each domain annotation database is set to `true`, or the `*_db` parameter paths (e.g., `pfam_db`, `funfam_db`) are set (i.e., not `null`), or the run is resumed after a successful database download, then the respective database will not be (re)downloaded. The full database links can be found in the main `nextflow.config` file, while minimal test versions can be found in the `test` and `test_full` profiles (i.e., `conf/test.config`, `conf/test_full.config`).
+
 [aria2](https://github.com/aria2/aria2/) is a lightweight multi-protocol & multi-source, cross platform download utility operated in command-line. It supports HTTP/HTTPS, FTP, SFTP, BitTorrent and Metalink.
 
 #### hmmer
@@ -86,10 +87,12 @@ The `seqkit` module is used for initial preprocessing (i.e., gap removal, conver
 - `domain_annotation/`
   - `pfam/`
     - `<samplename>.domtbl.gz`: `hmmer/hmmsearch` results along parameters info.
+  - `funfam/`
+    - `<samplename>.domtbl.gz`: `hmmer/hmmsearch` results along parameters info.
 
 </details>
 
-The `domain_annotation/pfam` folder contains a `.domtbl.gz` annotation file per input sample.
+Each of the `domain_annotation/` subfolders (e.g., `pfam`, `funfam`) contain a `.domtbl.gz` annotation file per input sample, depending on which domain annotation databases were used in the pipeline execution.
 
 [hmmer](https://github.com/EddyRivasLab/hmmer) is a fast and flexible alignment trimming tool that keeps phylogenetically informative sites and removes others.
 

main.nf

@@ -41,8 +41,11 @@ workflow NFCORE_PROTEINANNOTATOR {
         samplesheet,
         params.skip_preprocessing,
         params.skip_pfam,
-        params.pfam_latest_link,
         params.pfam_db,
+        params.pfam_latest_link,
+        params.skip_funfam,
+        params.funfam_db,
+        params.funfam_latest_link,
         params.skip_s4pred
     )
     emit:

nextflow.config

@@ -22,6 +22,9 @@ params {
     skip_pfam               = false
     pfam_db                 = null
     pfam_latest_link        = "https://ftp.ebi.ac.uk/pub/databases/Pfam/current_release/Pfam-A.hmm.gz"
+    skip_funfam             = false
+    funfam_db               = null
+    funfam_latest_link      = "https://download.cathdb.info/cath/releases/all-releases/v4_3_0/sequence-data/funfam-hmm3-v4_3_0.lib.gz"
     hmmsearch_evalue_cutoff = 0.001
 
     // InterProScan Options

nextflow_schema.json

@@ -286,6 +286,23 @@
                     "description": "InterPro hosted link to the latest Pfam HMM database file.",
                     "help_text": "Latest version should be a bit more than 350MB."
                 },
+                "skip_funfam": {
+                    "type": "boolean",
+                    "fa_icon": "fas fa-ban",
+                    "description": "Skip the domain annotation with the FunFam database.",
+                    "help": "Skips the domain annotation of input sequence against a FunFam database."
+                },
+                "funfam_db": {
+                    "type": "string",
+                    "format": "file-path",
+                    "description": "Path to an already installed FunFam HMM database (.lib.gz).",
+                    "help_text": "If left null and skip_funfam is false, the pipeline will start downloading the latest FunFam HMM library."
+                },
+                "funfam_latest_link": {
+                    "type": "string",
+                    "default": "https://download.cathdb.info/cath/releases/all-releases/v4_3_0/sequence-data/funfam-hmm3-v4_3_0.lib.gz",
+                    "description": "CATH hosted link to the latest available (v4_3_0) FunFam HMM database file."
+                },
                 "hmmsearch_evalue_cutoff": {
                     "type": "number",
                     "default": 0.001,

subworkflows/local/domain_annotation/main.nf

@@ -1,24 +1,31 @@
-include { ARIA2           } from '../../../modules/nf-core/aria2/main'
-include { HMMER_HMMSEARCH } from '../../../modules/nf-core/hmmer/hmmsearch/main'
+include { ARIA2 as ARIA2_PFAM                 } from '../../../modules/nf-core/aria2/main'
+include { ARIA2 as ARIA2_FUNFAM               } from '../../../modules/nf-core/aria2/main'
+include { HMMER_HMMSEARCH as HMMSEARCH_PFAM   } from '../../../modules/nf-core/hmmer/hmmsearch/main'
+include { HMMER_HMMSEARCH as HMMSEARCH_FUNFAM } from '../../../modules/nf-core/hmmer/hmmsearch/main'
 
 workflow DOMAIN_ANNOTATION {
     take:
-    ch_fasta          // channel: [ val(meta), [ fasta ] ]
-    skip_pfam         // boolean
-    pfam_latest_link  // string, path to the latest pfam HMM database, to download
-    pfam_db           // string, path to the pfam HMM database, if already exists
+    ch_fasta            // channel: [ val(meta), [ fasta ] ]
+    skip_pfam           // boolean
+    pfam_db             // string, path to the pfam HMM database, if already exists
+    pfam_latest_link    // string, path to the latest pfam HMM database, to download
+    skip_funfam         // boolean
+    funfam_db           // string, path to the funfam HMM database, if already exists
+    funfam_latest_link  // string, path to the latest funfam HMM database, to download
 
     main:
 
-    ch_versions = channel.empty()
+    ch_versions       = channel.empty()
+    ch_pfam_domains   = channel.empty()
+    ch_funfam_domains = channel.empty()
 
     if (!skip_pfam) {
         if (!pfam_db) {
             ch_pfam_link = channel.of([ [ id: 'pfam' ], pfam_latest_link ])
 
-            ARIA2( ch_pfam_link )
-            ch_versions = ch_versions.mix( ARIA2.out.versions )
-            ch_pfam_db = ARIA2.out.downloaded_file
+            ARIA2_PFAM( ch_pfam_link )
+            ch_versions = ch_versions.mix( ARIA2_PFAM.out.versions )
+            ch_pfam_db = ARIA2_PFAM.out.downloaded_file
         } else {
             ch_pfam_db = channel.of([ [ id: 'pfam' ], pfam_db ])
         }
@@ -27,11 +34,33 @@ workflow DOMAIN_ANNOTATION {
             .combine(ch_pfam_db)
             .map{ meta, seqs, _meta2, models -> [meta, models, seqs, false, false, true] }
 
-        HMMER_HMMSEARCH( ch_input_for_hmmsearch )
-        ch_versions = ch_versions.mix( HMMER_HMMSEARCH.out.versions.first() )
+        HMMSEARCH_PFAM( ch_input_for_hmmsearch )
+        ch_versions = ch_versions.mix( HMMSEARCH_PFAM.out.versions.first() )
+        ch_pfam_domains = HMMSEARCH_PFAM.out.domain_summary
+    }
+
+    if (!skip_funfam) {
+        if (!funfam_db) {
+            ch_funfam_link = channel.of([ [ id: 'funfam' ], funfam_latest_link ])
+
+            ARIA2_FUNFAM( ch_funfam_link )
+            ch_versions = ch_versions.mix( ARIA2_FUNFAM.out.versions )
+            ch_funfam_db = ARIA2_FUNFAM.out.downloaded_file
+        } else {
+            ch_funfam_db = channel.of([ [ id: 'funfam' ], funfam_db ])
+        }
+
+        ch_input_for_hmmsearch = ch_fasta
+            .combine(ch_funfam_db)
+            .map{ meta, seqs, _meta2, models -> [meta, models, seqs, false, false, true] }
+
+        HMMSEARCH_FUNFAM( ch_input_for_hmmsearch )
+        ch_versions = ch_versions.mix( HMMSEARCH_FUNFAM.out.versions.first() )
+        ch_funfam_domains = HMMSEARCH_FUNFAM.out.domain_summary
     }
 
     emit:
-    pfam_domains = HMMER_HMMSEARCH.out.domain_summary
-    versions     = ch_versions
+    pfam_domains   = ch_pfam_domains
+    funfam_domains = ch_funfam_domains
+    versions       = ch_versions
 }

subworkflows/local/domain_annotation/meta.yml

@@ -0,0 +1,60 @@
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/subworkflows/yaml-schema.json
+name: "domain_annotation"
+description: Annotate amino acid fasta files with selected HMM libraries such as Pfam and FunFam
+keywords:
+  - fasta
+  - sequences
+  - domain
+  - annotation
+  - database
+  - download
+  - HMM
+components:
+  - aria2
+  - hmmer/hmmsearch
+input:
+  - ch_fasta:
+      type: file
+      description: |
+        Amino acid fasta file containing amino acid sequences for annotation
+  - skip_pfam:
+      type: boolean
+      description: |
+        Skip domain annotation with Pfam
+  - pfam_db:
+      type: string
+      description: |
+        Path to an existing HMM Pfam library on the system. If provided, the ARIA2_PFAM db download will be skipped.
+  - pfam_latest_link:
+      type: string
+      description: |
+        Path to the latest Pfam HMM database, to download
+  - skip_funfam:
+      type: boolean
+      description: |
+        Skip domain annotation with FunFam
+  - funfam_db:
+      type: string
+      description: |
+        Path to an existing HMM FunFam library on the system. If provided, the ARIA2_FUNFAM db download will be skipped.
+  - funfam_latest_link:
+      type: string
+      description: |
+        Path to the latest FunFam HMM database, to download
+output:
+  - pfam_domains:
+      type: file
+      description: |
+        domtbl.gz files with pfam domain annotation for input amino acid sequences
+  - funfam_domains:
+      type: file
+      description: |
+        domtbl.gz files with funfam domain annotation for input amino acid sequences
+  - versions:
+      type: file
+      description: |
+        Versions file containing the software versions used in the workflow
+authors:
+  - "@vagkaratzas"
+maintainers:
+  - "@vagkaratzas"