Merge pull request #21705 from laraPPr/20241018160816_new_pr_CodingQu…

…arry20

{bio}[foss/2023a] funannotate v1.8.17 + required dependencies

easybuild/easyconfigs/c/CodingQuarry/CodingQuarry-2.0-foss-2023a.eb

@@ -0,0 +1,51 @@
+easyblock = 'MakeCp'
+
+name = 'CodingQuarry'
+version = '2.0'
+
+homepage = 'https://sourceforge.net/p/codingquarry'
+description = "Highly accurate hidden Markov model gene prediction in fungal genomes using RNA-seq transcripts"
+
+toolchain = {'name': 'foss', 'version': '2023a'}
+toolchainopts = {'openmp': True}
+
+source_urls = [SOURCEFORGE_SOURCE]
+sources = ['CodingQuarry_v%(version)s.tar.gz']
+patches = ['CodingQuarry-2.0_python3.patch']
+checksums = [
+    {'CodingQuarry_v2.0.tar.gz': '1198afbf7cebcf0975c5b20d92b7a2dd6d956072fcde6e86fdce6aeae4842504'},
+    {'CodingQuarry-2.0_python3.patch': '8e1b117431d8b104f2114875d8f751aa91c1c3c1b0ddd5a4f85251605c2ab9df'},
+]
+
+dependencies = [
+    ('Python', '3.11.3'),
+    ('Biopython', '1.83'),
+]
+
+buildopts = 'CFLAGS="$CFLAGS"'
+
+files_to_copy = [
+    (['CodingQuarry', 'CufflinksGTF_to_CodingQuarryGFF3.py'], 'bin'),
+    'QuarryFiles',
+    'TESTING',
+]
+
+fix_python_shebang_for = [
+    'bin/CufflinksGTF_to_CodingQuarryGFF3.py',
+    'QuarryFiles/scripts/*.py',
+]
+
+sanity_check_paths = {
+    'files': ['bin/CodingQuarry', 'bin/CufflinksGTF_to_CodingQuarryGFF3.py'],
+    'dirs': ['QuarryFiles/scripts', 'QuarryFiles/self_train', 'QuarryFiles/species', 'TESTING'],
+}
+
+sanity_check_commands = [
+    "CodingQuarry --help | grep '^CodingQuarry v. %(version)s'",
+    "mkdir -p %(builddir)s && cp -a %(installdir)s/TESTING %(builddir)s/TESTING",
+    "cd %(builddir)s/TESTING && CufflinksGTF_to_CodingQuarryGFF3.py Sp_transcripts.gtf > test.gff3",
+]
+
+modextravars = {'QUARRY_PATH': '%(installdir)s/QuarryFiles'}
+
+moduleclass = 'bio'

easybuild/easyconfigs/d/DBD-mysql/DBD-mysql-4.050-GCC-12.3.0.eb

@@ -0,0 +1,30 @@
+easyblock = 'PerlModule'
+
+name = 'DBD-mysql'
+version = '4.050'
+
+homepage = 'https://metacpan.org/pod/distribution/DBD-mysql/lib/DBD/mysql.pm'
+description = "Perl binding for MySQL"
+
+toolchain = {'name': 'GCC', 'version': '12.3.0'}
+
+source_urls = ['https://cpan.metacpan.org/authors/id/D/DV/DVEEDEN']
+sources = [SOURCE_TAR_GZ]
+checksums = ['4f48541ff15a0a7405f76adc10f81627c33996fbf56c95c26c094444c0928d78']
+
+dependencies = [
+    ('Perl', '5.36.1'),
+    ('Perl-bundle-CPAN', '5.36.1'),
+    ('MariaDB', '11.6.0'),
+    ('zlib', '1.2.13'),
+    ('OpenSSL', '1.1', '', SYSTEM),
+]
+
+options = {'modulename': 'DBD::mysql'}
+
+sanity_check_paths = {
+    'files': ['lib/perl5/site_perl/%%(perlver)s/%s-linux-thread-multi/DBD/mysql.pm' % ARCH],
+    'dirs': ['lib/perl5/site_perl/%%(perlver)s/%s-linux-thread-multi/DBD/mysql' % ARCH],
+}
+
+moduleclass = 'data'

easybuild/easyconfigs/e/eggnog-mapper/eggnog-mapper-2.1.12-foss-2023a.eb

@@ -0,0 +1,56 @@
+# Eggnog DB installation instructions:
+# 1. 'export EGGNOG_DATA_DIR=/<storage-location>/eggnog-mapper-data'
+# 2. run 'download_eggnog_data.py'
+# 3. Check the expected DB version with 'emapper.py --version'
+
+easyblock = 'PythonPackage'
+
+name = 'eggnog-mapper'
+version = '2.1.12'
+
+homepage = 'https://github.com/eggnogdb/eggnog-mapper'
+description = """EggNOG-mapper is a tool for fast functional annotation of novel
+sequences. It uses precomputed orthologous groups and phylogenies from the
+eggNOG database (http://eggnog5.embl.de) to transfer functional information from
+fine-grained orthologs only. Common uses of eggNOG-mapper include the annotation
+of novel genomes, transcriptomes or even metagenomic gene catalogs."""
+
+toolchain = {'name': 'foss', 'version': '2023a'}
+
+github_account = 'eggnogdb'
+source_urls = [GITHUB_SOURCE]
+sources = ['%(version)s.tar.gz']
+checksums = ['b3c53fb0e606a5cfec75cbc84f7c215f57f43ce00d8e50f449513acdad76da73']
+
+dependencies = [
+    ('Python', '3.11.3'),
+    ('Biopython', '1.83'),
+    ('HMMER', '3.4'),
+    ('DIAMOND', '2.1.8'),
+    ('prodigal', '2.6.3'),
+    ('wget', '1.24.5'),
+    ('MMseqs2', '14-7e284'),
+    ('XlsxWriter', '3.1.3'),
+]
+
+# strip out (too) strict version requirements for dependencies
+preinstallopts = "sed -i 's/==[0-9.]*//g' setup.cfg && "
+
+use_pip = True
+sanity_pip_check = True
+download_dep_fail = True
+
+sanity_check_paths = {
+    'files': ['bin/create_dbs.py', 'bin/download_eggnog_data.py', 'bin/emapper.py'],
+    'dirs': ['lib/python%(pyshortver)s/site-packages'],
+}
+
+sanity_check_commands = [
+    'download_eggnog_data.py --help',
+    'create_dbs.py --help',
+    'emapper.py --version | grep %(version)s',
+]
+
+options = {'modulename': 'eggnogmapper'}
+
+moduleclass = 'bio'

easybuild/easyconfigs/f/funannotate/funannotate-1.8.17-foss-2023a.eb

@@ -0,0 +1,94 @@
+easyblock = 'PythonBundle'
+
+name = 'funannotate'
+version = '1.8.17'
+
+homepage = 'https://funannotate.readthedocs.io'
+description = """funannotate is a pipeline for genome annotation (built specifically for fungi, but will also work
+ with higher eukaryotes)"""
+
+toolchain = {'name': 'foss', 'version': '2023a'}
+
+# see also https://github.com/nextgenusfs/funannotate/blob/master/docs/dependencies.rst
+dependencies = [
+    ('Python', '3.11.3'),
+    ('SciPy-bundle', '2023.07'),
+    ('Biopython', '1.83'),
+    ('GOATOOLS', '1.4.5'),
+    ('matplotlib', '3.7.2'),
+    ('scikit-learn', '1.3.1'),
+    ('Seaborn', '0.13.2'),
+    ('tbl2asn', '20230713'),
+    ('DBD-mysql', '4.050'),
+    ('CodingQuarry', '2.0'),
+    ('Trinity', '2.15.2'),
+    ('AUGUSTUS', '3.5.0'),
+    ('BamTools', '2.5.2'),
+    ('BEDTools', '2.31.0'),
+    ('BLAST+', '2.14.1'),  # provides makeblastdb, tblastn
+    ('BLAT', '3.7'),
+    ('DIAMOND', '2.1.8'),
+    ('eggnog-mapper', '2.1.12'),
+    ('ETE', '3.1.3'),
+    ('Exonerate', '2.4.0'),
+    ('FASTA', '36.3.8i'),
+    ('GlimmerHMM', '3.0.4c'),
+    ('GMAP-GSNAP', '2023-04-20'),  # provides gmap
+    ('GeneMark-ET', '4.72'),  # provides gmes_petap.pl
+    ('HISAT2', '2.2.1'),
+    ('HMMER', '3.4'),  # provides hmmscan, hmmsearch
+    ('kallisto', '0.51.1'),
+    ('MAFFT', '7.520', '-with-extensions'),
+    ('minimap2', '2.26'),
+    ('pigz', '2.8'),
+    ('Proteinortho', '6.3.2'),
+    ('Kent_tools', '468'),  # provides pslCDnaFilter
+    ('Salmon', '1.10.3'),
+    ('SAMtools', '1.18'),
+    ('SignalP', '6.0h', '-fast'),
+    ('SNAP-HMM', '20221022'),
+    ('StringTie', '2.2.3'),
+    ('tRNAscan-SE', '2.0.12'),
+    ('tantan', '50'),
+    ('trimAl', '1.4.1'),
+    ('Trimmomatic', '0.39', '-Java-11', SYSTEM),
+    ('BioPerl', '1.7.8'),
+    ('EVidenceModeler', '2.1.0'),
+]
+
+use_pip = True
+
+exts_list = [
+    ('distro', '1.9.0', {
+        'checksums': ['2fa77c6fd8940f116ee1d6b94a2f90b13b5ea8d019b98bc8bafdcabcdd9bdbed'],
+    }),
+    ('natsort', '8.4.0', {
+        'checksums': ['45312c4a0e5507593da193dedd04abb1469253b601ecaf63445ad80f0a1ea581'],
+    }),
+    (name, version, {
+        'checksums': ['bdadfd7a5636383c1c40c26dab37c5908a77e8c4064adced84f1ba9e86187a04'],
+        'preinstallopts': (
+            """sed -i 's|REQUIRES_PYTHON = ">=3.6.0, <3.10"|REQUIRES_PYTHON = ">=3.6.0"|' setup.py && """
+            """sed -i 's|"biopython<1.80",|"biopython",|' setup.py && """
+        )
+    }),
+]
+
+sanity_check_paths = {
+    'files': ['bin/funannotate'],
+    'dirs': [],
+}
+
+modextrapaths = {
+    'GENEMARK_PATH': '$EBROOTGENEMARKMINET',
+}
+
+sanity_check_commands = [
+    "funannotate --help 2>&1 | grep 'Usage:[ ]*funannotate'",
+    "funannotate check --show-versions",
+]
+
+
+sanity_pip_check = True
+
+moduleclass = 'bio'

easybuild/easyconfigs/g/GATK/GATK-4.3.0.0-GCCcore-12.3.0-Java-11.eb

@@ -0,0 +1,55 @@
+##
+# This file is an EasyBuild reciPY as per https://github.com/easybuilders/easybuild
+#
+# Copyright:: Copyright 2012-2013 Cyprus Institute / CaSToRC, University of Luxembourg / LCSB
+# Authors::   George Tsouloupas <g.tsouloupas@cyi.ac.cy>, Fotis Georgatos <fotis.georgatos@uni.lu>,
+#             Kenneth Hoste (UGent)
+# License::   MIT/GPL
+# $Id$
+#
+# This work implements a part of the HPCBIOS project and is a component of the policy:
+# http://hpcbios.readthedocs.org/en/latest/HPCBIOS_2012-94.html
+# Modified by: Adam Huffman, Jonas Demeulemeester
+# The Francis Crick Institute
+# Modified for version 4.0.5.1 by: Ruben van Dijk, University of Groningen
+# Modified for version 4.2.3.0 by: J. Sassmannshausen / GSTT
+# Modified for version 4.4.0.0 by: Thomas Eylenbosch / Gluo NV
+##
+
+easyblock = 'Tarball'
+
+name = 'GATK'
+version = '4.3.0.0'
+versionsuffix = '-Java-%(javaver)s'
+
+homepage = 'https://www.broadinstitute.org/gatk/'
+description = """The Genome Analysis Toolkit or GATK is a software package developed at the Broad Institute
+ to analyse next-generation resequencing data. The toolkit offers a wide variety of tools,
+ with a primary focus on variant discovery and genotyping as well as strong emphasis on
+ data quality assurance. Its robust architecture, powerful processing engine and
+ high-performance computing features make it capable of taking on projects of any size."""
+
+toolchain = {'name': 'GCCcore', 'version': '12.3.0'}
+
+source_urls = ['https://github.com/broadinstitute/gatk/releases/download/%(version)s/']
+sources = ['gatk-%(version)s.zip']
+checksums = ['e2c27229b34c3e22445964adf00639a0909887bbfcc040f6910079177bc6e2dd']
+
+dependencies = [
+    ('Java', '11', '', SYSTEM),
+    ('Python', '3.11.3'),
+]
+
+modextrapaths = {'PATH': ''}
+
+sanity_check_paths = {
+    'files': ['gatk'],
+    'dirs': [],
+}
+
+sanity_check_commands = [
+    "gatk --help",
+    "gatk --list",
+]
+
+moduleclass = 'bio'

easybuild/easyconfigs/g/GlimmerHMM/GlimmerHMM-3.0.4c-GCC-12.3.0.eb

@@ -0,0 +1,57 @@
+# This file is an EasyBuild reciPY as per https://github.com/easybuilders/easybuild
+
+easyblock = 'MakeCp'
+
+name = 'GlimmerHMM'
+version = '3.0.4c'
+
+homepage = 'https://ccb.jhu.edu/software/glimmerhmm'
+description = """GlimmerHMM is a new gene finder based on a Generalized Hidden Markov Model.
+ Although the gene finder conforms to the overall mathematical framework of a GHMM, additionally
+ it incorporates splice site models adapted from the GeneSplicer program and a decision tree adapted
+ from GlimmerM. It also utilizes Interpolated Markov Models for the coding and noncoding models."""
+
+toolchain = {'name': 'GCC', 'version': '12.3.0'}
+toolchainopts = {'pic': True}
+
+source_urls = ['https://ccb.jhu.edu/software/%(namelower)s/dl']
+sources = [SOURCE_TAR_GZ]
+checksums = ['31ee2ceb8f31338205b2de626d83d0f92d2cd55a04d48a6803193a2d0ad1b4a3']
+
+dependencies = [
+    ('Perl', '5.36.1'),
+]
+
+start_dir = 'sources'
+
+# make sure -O0 is not used as compiler option
+prebuildopts = "ls makefile train/makefile | xargs sed -i 's/-O0 .*//g' && "
+
+# also build in 'train' subdirectory to overwrite pre-compiled binaries
+buildopts = "&& cd ../train && make"
+
+local_train_files = ['build1', 'build2', 'build-icm', 'build-icm-noframe', 'erfapp', 'falsecomp',
+                     'findsites', 'karlin', 'score', 'score2', 'scoreATG', 'scoreATG2', 'scoreSTOP',
+                     'scoreSTOP2', 'splicescore', 'trainGlimmerHMM']
+files_to_copy = [
+    (['sources/%(namelower)s'], 'bin'),
+    (['train/%s' % x for x in local_train_files], 'bin'),
+    (['train/*.pm'], 'lib/perl%(perlmajver)s'),
+    'trained_dir', 'README', 'train/readme.train',
+]
+
+fix_perl_shebang_for = ['bin/trainGlimmerHMM']
+
+sanity_check_paths = {
+    'files': ['bin/%(namelower)s'],
+    'dirs': ['trained_dir'],
+}
+
+sanity_check_commands = [
+    "%(namelower)s -h",
+    r"trainGlimmerHMM -h 2>&1 | grep '^[ ]*Train GlimmerHMM module'",
+]
+
+modextrapaths = {'PERL5LIB': 'lib/perl%(perlmajver)s'}
+
+moduleclass = 'bio'

easybuild/easyconfigs/i/Infernal/Infernal-1.1.5-foss-2023a.eb

@@ -0,0 +1,39 @@
+##
+# This file is an EasyBuild reciPY as per https://github.com/easybuilders/easybuild
+#
+# Copyright:: Copyright 2012-2014 Uni.Lu/LCSB, NTUA
+# Authors::   Cedric Laczny <cedric.laczny@uni.lu>, Fotis Georgatos <fotis@cern.ch>
+# License::   MIT/GPL
+# Updated:: Denis Kristak (INUITS)
+# $Id$
+#
+# This work implements a part of the HPCBIOS project and is a component of the policy:
+# http://hpcbios.readthedocs.org/en/latest/HPCBIOS_2012-94.html
+##
+
+easyblock = 'ConfigureMake'
+
+name = 'Infernal'
+version = "1.1.5"
+
+homepage = 'http://eddylab.org/infernal/'
+description = """Infernal ("INFERence of RNA ALignment") is for searching DNA sequence databases
+ for RNA structure and sequence similarities."""
+
+toolchain = {'name': 'foss', 'version': '2023a'}
+toolchainopts = {'pic': True}
+
+source_urls = ['http://eddylab.org/%(namelower)s']
+sources = [SOURCELOWER_TAR_GZ]
+checksums = ['ad4ddae02f924ca7c85bc8c4a79c9f875af8df96aeb726702fa985cbe752497f']
+
+local_bins = ['align', 'build', 'calibrate', 'convert', 'emit', 'fetch', 'press', 'scan', 'search', 'stat']
+
+sanity_check_paths = {
+    'files': ['bin/cm%s' % x for x in local_bins],
+    'dirs': []
+}
+
+sanity_check_commands = ['cm%s -h' % x for x in local_bins]
+
+moduleclass = 'bio'