Merge pull request #217 from maguire-lab/master

Update of pre-installed db, fix abricate-getdb script, and transition CI to github actions

Author: tseemann

.github/workflows/ci.yml

@@ -0,0 +1,60 @@
+name: CI
+
+on:
+  # Triggers the workflow on push or pull request events but only for the "master" branch
+  push:
+    branches: [ "master" ]
+  pull_request:
+    branches: [ "master" ]
+
+  # Allows you to run this workflow manually from the Actions tab
+  workflow_dispatch:
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      # Checks-out your repository under $GITHUB_WORKSPACE, so your job can access it
+      - uses: actions/checkout@v4
+      - uses: shogo82148/actions-setup-perl@v1
+        with:
+          perl-version: "5.38"
+
+      - name: Install
+        run: |
+          sudo apt-get install -y zip gzip git curl libdb-dev
+          cpanm --quiet --notest JSON Path::Tiny LWP::Simple Bio::Perl
+          wget http://ftp.ncbi.nlm.nih.gov/blast/executables/blast+/2.16.0/ncbi-blast-2.16.0+-x64-linux.tar.gz
+          tar zxvf ncbi-blast-2.16.0+-x64-linux.tar.gz
+          curl https://raw.githubusercontent.com/tseemann/any2fasta/master/any2fasta > bin/any2fasta
+          chmod +x bin/any2fasta
+          export PATH=$PWD/bin:$PWD/ncbi-blast-2.16.0+/bin:$PATH
+
+      - name: Test
+        run: |
+          export PATH=$PWD/bin:$PWD/ncbi-blast-2.16.0+/bin:$PATH
+          abricate --setupdb
+          abricate --version
+          abricate --help
+          abricate --check
+          abricate --list
+          ! abricate --doesnotexist || echo "fail expected"
+          ! abricate --threads 0
+          ! (abricate test/assembly.fa | grep '~~~')
+          abricate test/assembly.fa > 1.tab
+          abricate test/assembly.fa.gz > 2.tab
+          abricate test/assembly.gbk > 3.tab
+          abricate test/assembly.gbk.gz > 4.tab
+          abricate --nopath test/assembly.gbk.gz | grep '^assembly.gbk.gz'
+          abricate --summary {1,2,3,4}.tab > summary.tab
+          abricate --summary 1.tab 2.tab 1.tab |& grep 'duplicate'
+          abricate --summary <(cat 1.tab 2.tab 3.tab) | wc -l | grep -w 4
+          abricate test/assembly.txt |& grep ERROR
+          abricate not_exist.embl |& grep ERROR
+          abricate --threads `nproc` test/assembly.fa.bz2 | grep -i FOSFOMYCIN
+          abricate --threads `nproc` test/assembly.fa.bz2 | grep -i lactam
+          for DB in `abricate --list | cut -f1 | tail -n +2`; do abricate --db $DB test/assembly.fa > /dev/null ; done
+          abricate-get_db --help
+          abricate-get_db --db ncbi --dbdir .
+          ! grep 'FUSIDIC ACID' ncbi/sequences
+          abricate --threads `nproc` --fofn test/fofn.txt

README.md

@@ -1,4 +1,3 @@
-[![Build Status](https://travis-ci.org/tseemann/abricate.svg?branch=master)](https://travis-ci.org/tseemann/abricate) 
 [![License: GPL v2](https://img.shields.io/badge/License-GPL%20v2-blue.svg)](https://www.gnu.org/licenses/old-licenses/gpl-2.0.en.html)
 [![European Galaxy server](https://img.shields.io/badge/usegalaxy-.eu-brightgreen?logo=data:image/png;base64,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)](https://usegalaxy.eu/root?tool_id=abricate)
 ![Don't judge me](https://img.shields.io/badge/Language-Perl_5-steelblue.svg)
@@ -158,16 +157,19 @@ columns:
 ```
 % abricate --list
 
-DATABASE       SEQUENCES  DBTYPE  DATE
-argannot       1749       nucl    2019-Jul-28
-card           2241       nucl    2019-Jul-28
-ecoh           597        nucl    2019-Jul-28
-ecoli_vf       2701       nucl    2019-Jul-28
-megares        6635       nucl    2020-Feb-20
-ncbi           4324       nucl    2019-Jul-28
-plasmidfinder  263        nucl    2019-Jul-28
-resfinder      2434       nucl    2019-Jul-28
-vfdb           2597       nucl    2019-Jul-28
+DATABASE	SEQUENCES	DBTYPE	DATE
+ecoh	597	nucl	2025-Apr-14
+bacmet2	746	nucl	2025-Apr-14
+ncbi	7456	nucl	2025-Apr-14
+victors	4545	nucl	2025-Apr-14
+vfdb	4394	nucl	2025-Apr-14
+megares	7424	nucl	2025-Apr-14
+card	6048	nucl	2025-Apr-14
+plasmidfinder	488	nucl	2025-Apr-14
+resfinder	3174	nucl	2025-Apr-14
+argannot	2224	nucl	2025-Apr-14
+ecoli_vf	2701	nucl	2025-Apr-14
+
 ```
 
 The default database is `ncbi`.
@@ -257,14 +259,14 @@ If you publish the results of Abricate please cite both the software _and_
 the appropriate database you used with `--db`
 
 * Seemann T, *Abricate*, **Github** `https://github.com/tseemann/abricate`
-* NCBI AMRFinderPlus - [doi: 10.1128/AAC.00483-19](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6811410)
-* CARD - [doi:10.1093/nar/gkw1004](https://www.ncbi.nlm.nih.gov/pubmed/27789705)
-* Resfinder - [doi:10.1093/jac/dks261](https://www.ncbi.nlm.nih.gov/pubmed/22782487)
+* NCBI AMRFinderPlus - [doi:10.1038/s41598-021-91456-0](https://pubmed.ncbi.nlm.nih.gov/34135355/)
+* CARD - [doi:10.1093/nar/gkac920](https://pubmed.ncbi.nlm.nih.gov/36263822/)
+* Resfinder - [doi:10.1093/jac/dkaa345](https://pubmed.ncbi.nlm.nih.gov/32780112/)
 * ARG-ANNOT - [doi:10.1128/AAC.01310-13](https://www.ncbi.nlm.nih.gov/pubmed/24145532)
 * VFDB - [doi:10.1093/nar/gkv1239](https://www.ncbi.nlm.nih.gov/pubmed/26578559)
 * PlasmidFinder - [doi:10.1128/AAC.02412-14](https://www.ncbi.nlm.nih.gov/pubmed/24777092)
 * EcOH - [doi:10.1099/mgen.0.000064](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5343136/)
-* MEGARES 2.00 - [doi:10.1093/nar/gkz1010](https://academic.oup.com/nar/article/48/D1/D561/5624973)
+* MEGARES 3.00 - [doi:10.1093/nar/gkac1047](https://pubmed.ncbi.nlm.nih.gov/36382407/)
 ## Issues
 
 Please report problems to the [Issues Page](https://github.com/tseemann/abricate/issues).

bin/abricate-get_db

@@ -246,7 +246,7 @@ sub get_tag {
 sub get_ncbi {
   my $AFP = "https://ftp.ncbi.nlm.nih.gov/pathogen/Antimicrobial_resistance/AMRFinderPlus/database/latest";
   #my $src = "https://ftp.ncbi.nlm.nih.gov/pathogen/Antimicrobial_resistance/AMRFinderPlus/data/latest/AMR_CDS";
-  my $src = "$AFP/AMR_CDS";
+  my $src = "$AFP/AMR_CDS.fa";
   my $name = "amr_cds.ffn";
   #my $src2 = "https://ftp.ncbi.nlm.nih.gov/pathogen/Antimicrobial_resistance/AMRFinderPlus/data/latest/ReferenceGeneCatalog.txt";
   my $src2 = "$AFP/ReferenceGeneCatalog.txt";
@@ -347,7 +347,7 @@ sub get_plasmidfinder {
 #..............................................................................
 sub get_megares {
   my $zip = "megares.zip";
-  download('https://megares.meglab.org/download/megares_v2.00.zip', $zip);
+  download('https://www.meglab.org/downloads/megares_v3.00.zip', $zip);
   system("unzip -j -u $zip");
   my $seqs = load_fasta( glob("megares_drugs_*.fasta") );
   my @okseq;
@@ -380,7 +380,7 @@ sub get_argannot {
   download(
 #    'http://www.mediterranee-infection.com/arkotheque/client/ihumed/_depot_arko/articles/1425/argannot-aa-v3-march2017_doc.fasta',
 #    'http://www.mediterranee-infection.com/arkotheque/client/ihumed/_depot_arko/articles/691/argannot-nt_doc.fasta',
-    'https://www.mediterranee-infection.com/wp-content/uploads/2019/06/ARG_ANNOT_V5_Nt_JUNE2019.txt',
+    'https://www.mediterranee-infection.com/wp-content/uploads/2019/09/ARG-ANNOT_NT_V6_July2019.txt',
     $fasta
   );
 
@@ -501,10 +501,10 @@ sub get_victors {
   # the PROT data is in .faa and has the protein ref and /product
 
   #>gi|115534241:2616-3152 Campylobacter jejuni plasmid pCJ01, complete sequence
-  download('http://www.phidias.us/victors/downloads/gen_downloads.php', 'victors.ffn');
+  download('http://phidias.us/victors/downloads/gen_downloads.php', 'victors.ffn');
 
   #>gi|115534244|ref|YP_783826.1| hypothetical protein pCJ01p4 [Campylobacter jejuni]
-  download('http://www.phidias.us/victors/downloads/gen_downloads_protein.php', 'victors.faa');
+  download('http://phidias.us/victors/downloads/gen_downloads_protein.php', 'victors.faa');
 
   my %gi;
   open my $FAA, '<', 'victors.faa';