BUSCO version 5.8.3

config/easybuild/easyconfigs/a/AUGUSTUS/AUGUSTUS-3.5.0-foss-2023b.eb

@@ -0,0 +1,79 @@
+# Updated by: Pavel Grochal (INUITS)
+# License: GPLv2
+
+easyblock = 'ConfigureMake'
+
+name = 'AUGUSTUS'
+version = '3.5.0'
+
+homepage = 'https://bioinf.uni-greifswald.de/augustus/'
+description = "AUGUSTUS is a program that predicts genes in eukaryotic genomic sequences"
+
+toolchain = {'name': 'foss', 'version': '2023b'}
+
+github_account = 'Gaius-Augustus'
+source_urls = [GITHUB_SOURCE]
+sources = ['v%(version)s.tar.gz']
+patches = ['%(name)s-3.5.0_sqliteDB_cstdint_include.patch']
+checksums = [
+    '5ed6ce6106303b800c5e91d37a250baff43b20824657b853ae04d11ad8bdd686',  # v3.5.0.tar.gz
+    '1c5e1813f100531ec8a9fd0bc948612d37a5676eb8f5d59601ed2cfed65fa255',  # AUGUSTUS-3.5.0_sqliteDB_cstdint_include.patch
+]
+
+builddependencies = [
+    ('Python', '3.11.5'),
+]
+
+dependencies = [
+    ('zlib', '1.2.13'),
+    ('Boost', '1.83.0'),
+    ('GSL', '2.7'),
+    ('SAMtools', '1.19.2'),
+    ('HTSlib', '1.19.1'),  # also provides tabix
+    ('BCFtools', '1.19'),
+    #('lpsolve', '5.6.20'),  # lpsolve from r-bundle-cran
+    ('R-bundle-CRAN', '2024.06'),
+    ('SuiteSparse', '7.7.0', '-METIS-5.1.0'),
+    ('BamTools', '2.5.2'),
+    ('SQLite', '3.43.1'),
+    ('lpsolve', '5.5.2.11'),
+]
+
+skipsteps = ['configure']
+
+# run "make clean" to avoid using binaries included with the source tarball
+prebuildopts = "make clean && "
+
+_tmpl = 'INCLUDE_PATH_{dep}=-I{root}{incl} LIBRARY_PATH_{dep}="-L{root}{lib} -Wl,-rpath,{root}{lib}"'
+
+buildopts = ' '.join([
+    'COMPGENEPRED=true SQLITE=true ZIPINPUT=true MYSQL=false CXX="$CXX" ',
+    _tmpl.format(dep='ZLIB', root='$EBROOTZLIB', incl='/include', lib='/lib'),
+    _tmpl.format(dep='BOOST', root='$EBROOTBOOST', incl='/include', lib='/lib'),
+    _tmpl.format(dep='LPSOLVE', root='$EBROOTLPSOLVE', incl='/include', lib='/lib'),
+    _tmpl.format(dep='SUITESPARSE', root='$EBROOTSUITESPARSE', incl='/include', lib='/lib'),
+    _tmpl.format(dep='GSL', root='$EBROOTGSL', incl='/include', lib='/lib'),
+    _tmpl.format(dep='SQLITE', root='$EBROOTSQLITE', incl='/include', lib='/lib'),
+    _tmpl.format(dep='BAMTOOLS', root='$EBROOTBAMTOOLS', incl='/include/bamtools', lib='/lib'),
+    _tmpl.format(dep='HTSLIB', root='$EBROOTHTSLIB', incl='/include/htslib', lib='/lib'),
+])
+
+preinstallopts = "sed -i '/ln -sf/d' Makefile && "
+installopts = 'INSTALLDIR=%(installdir)s '
+
+sanity_check_paths = {
+    'files': ['bin/augustus', 'bin/bam2hints', 'bin/etraining', 'bin/fastBlockSearch',
+              'bin/filterBam', 'bin/getSeq', 'bin/homGeneMapping', 'bin/joingenes',
+              'bin/load2sqlitedb', 'bin/prepareAlign'],
+    'dirs': ['config', 'scripts'],
+}
+sanity_check_commands = ['augustus --help']
+
+modextrapaths = {'PATH': 'scripts'}
+modextravars = {
+    'AUGUSTUS_BIN_PATH': '%(installdir)s/bin',
+    'AUGUSTUS_CONFIG_PATH': '%(installdir)s/config',
+    'AUGUSTUS_SCRIPTS_PATH': '%(installdir)s/scripts',
+}
+
+moduleclass = 'bio'

config/easybuild/easyconfigs/a/AUGUSTUS/AUGUSTUS-3.5.0_sqliteDB_cstdint_include.patch

@@ -0,0 +1,43 @@
+From 1ed97dc4ce2909c5f89737005b8ea4a664fbe728 Mon Sep 17 00:00:00 2001
+From: Kuoi <[email protected]>
+Date: Sun, 11 Jun 2023 23:47:49 +0800
+Subject: [PATCH] fix: gcc13 failed with this
+
+---
+ include/sqliteDB.hh | 1 +
+ 1 file changed, 1 insertion(+)
+
+diff --git a/include/sqliteDB.hh b/include/sqliteDB.hh
+index e66374b9d..67a37826f 100644
+--- a/include/sqliteDB.hh
++++ b/include/sqliteDB.hh
+@@ -11,6 +11,7 @@
+ #include <string>
+ #include <vector>
+ #include <sqlite3.h>
++#include <cstdint>
+
+ using namespace std;
+
+From 3dbe752e4cf3f6778168166a2c662d02d8623f15 Mon Sep 17 00:00:00 2001
+From: Kuoi <[email protected]>
+Date: Mon, 12 Jun 2023 01:32:02 +0800
+Subject: [PATCH] fix: without it compile fail
+
+---
+ auxprogs/homGeneMapping/include/sqliteDB.hh | 1 +
+ 1 file changed, 1 insertion(+)
+
+diff --git a/auxprogs/homGeneMapping/include/sqliteDB.hh b/auxprogs/homGeneMapping/include/sqliteDB.hh
+index e842bcf9d..67338bef4 100644
+--- a/auxprogs/homGeneMapping/include/sqliteDB.hh
++++ b/auxprogs/homGeneMapping/include/sqliteDB.hh
+@@ -13,6 +13,7 @@
+ #include <string>
+ #include <vector>
+ #include <sqlite3.h>
++#include <cstdint>
+
+ using namespace std;
+
+

config/easybuild/easyconfigs/b/BBMap/BBMap-39.01-GCC-13.2.0.eb

@@ -0,0 +1,34 @@
+easyblock = 'MakeCp'
+
+name = 'BBMap'
+version = '39.01'
+
+homepage = 'https://sourceforge.net/projects/bbmap/'
+description = """BBMap short read aligner, and other bioinformatic tools."""
+
+toolchain = {'name': 'GCC', 'version': '13.2.0'}
+
+source_urls = [SOURCEFORGE_SOURCE]
+sources = ['%(name)s_%(version)s.tar.gz']
+checksums = ['98608da50130c47f3abd095b889cc87f60beeb8b96169b664bc9d849abe093e6']
+
+dependencies = [('Java', '11.0.20', '', SYSTEM)]
+
+prebuildopts = 'cd jni && '
+
+local_suff = {'Darwin': 'osx', 'Linux': 'linux'}[OS_TYPE]
+buildopts = "-f makefile.%s" % local_suff
+
+files_to_copy = ['*']
+
+sanity_check_paths = {
+    'files': ['bbmap.sh', 'jni/libbbtoolsjni.%s' % SHLIB_EXT],
+    'dirs': []
+}
+
+modextrapaths = {'PATH': ''}
+
+modloadmsg = "For improved speed, add 'usejni=t' to the command line of %(name)s tools which support the use of the"
+modloadmsg += " compiled jni C code.\n"
+
+moduleclass = 'bio'

config/easybuild/easyconfigs/b/BLAST+/BLAST+-2.14.1-gompi-2023b.eb

@@ -0,0 +1,56 @@
+# #
+# EasyBuild reciPY as per https://github.com/easybuilders/easybuild
+#
+# Copyright:: Copyright 2012-2014 Uni.Lu/LCSB, NTUA
+# Authors::   Fotis Georgatos <[email protected]>, Kenneth Hoste (UGent)
+# License::   MIT/GPL
+# $Id$
+#
+# This work implements a part of the HPCBIOS project and is a component of
+# the policy: https://hpcbios.readthedocs.org/en/latest/HPCBIOS_2012-94.html
+# #
+
+easyblock = 'ConfigureMake'
+
+name = 'BLAST+'
+version = '2.14.1'
+
+homepage = 'https://blast.ncbi.nlm.nih.gov/'
+description = """Basic Local Alignment Search Tool, or BLAST, is an algorithm
+ for comparing primary biological sequence information, such as the amino-acid
+ sequences of different proteins or the nucleotides of DNA sequences."""
+
+toolchain = {'name': 'gompi', 'version': '2023b'}
+toolchainopts = {'usempi': True, 'pic': True}
+
+source_urls = ['https://ftp.ncbi.nlm.nih.gov/blast/executables/%(namelower)s/%(version)s/']
+sources = ['ncbi-blast-%(version)s+-src.tar.gz']
+checksums = ['712c2dbdf0fb13cc1c2d4f4ef5dd1ce4b06c3b57e96dfea8f23e6e99f5b1650e']
+
+builddependencies = [('cpio', '2.15')]
+
+dependencies = [
+    ('zlib', '1.2.13'),
+    ('bzip2', '1.0.8'),
+    ('PCRE', '8.45'),
+    ('Boost', '1.83.0'),
+    ('GMP', '6.3.0'),
+    ('libpng', '1.6.40'),
+    ('libjpeg-turbo', '3.0.1'),
+    ('LMDB', '0.9.31'),
+]
+
+# remove line that prepends system paths to $PATH from configure script
+preconfigopts = r'sed -i "s|^PATH=\(.*\)$|#PATH=\1 |" %(start_dir)s/src/build-system/configure && '
+
+configopts = "--with-z=$EBROOTZLIB --with-bz2=$EBROOTBZIP2 "
+configopts += "--with-pcre=$EBROOTPCRE --with-boost=$EBROOTBOOST "
+configopts += "--with-gmp=$EBROOTGMP --with-png=$EBROOTLIBPNG "
+configopts += "--with-jpeg=$EBROOTLIBJPEGMINTURBO --with-lmdb=$EBROOTLMDB"
+
+sanity_check_paths = {
+    'files': ['bin/blastn', 'bin/blastp', 'bin/blastx'],
+    'dirs': [],
+}
+
+moduleclass = 'bio'

config/easybuild/easyconfigs/b/BUSCO/BUSCO-5.8.3-foss-2023b.eb

@@ -0,0 +1,57 @@
+# Updated by: Pavel Grochal (INUITS)
+# Updated by: Sebastien Moretti (SIB)
+
+easyblock = 'PythonPackage'
+
+name = 'BUSCO'
+version = '5.8.3'
+
+homepage = 'https://busco.ezlab.org/'
+description = "BUSCO: assessing genome assembly and annotation completeness with single-copy orthologs"
+# software_license = 'LicenseMIT'
+
+toolchain = {'name': 'foss', 'version': '2023b'}
+
+source_urls = ['https://gitlab.com/ezlab/%(namelower)s/-/archive/%(version)s']
+sources = [SOURCELOWER_TAR_BZ2]
+checksums = ['b1a46c93b636ca59fd7529f8fba28774d9998c9a03d3f1fa8776a1fb08aa8b3f']
+
+dependencies = [
+    ('Python', '3.11.5'),
+    ('SciPy-bundle', '2023.11'),  # for pandas
+    ('R', '4.4.1'),  # for R and ggplot2
+    ('Biopython', '1.84'),
+    ('BLAST+', '2.14.1'),
+    ('HMMER', '3.4'),
+    ('prodigal', '2.6.3'),
+    ('AUGUSTUS', '3.5.0'),
+    ('SEPP', '4.5.5'),
+    ('MetaEuk', '7'),
+    ('BBMap', '39.01'),
+    ('miniprot', '0.14'),
+]
+
+use_pip = True
+download_dep_fail = True
+sanity_pip_check = True
+
+postinstallcmds = [
+    'mkdir -p %(installdir)s/bin %(installdir)s/doc',
+    'cp %(builddir)s/%(namelower)s-%(version)s/scripts/* %(installdir)s/bin',
+    'cp %(builddir)s/%(namelower)s-%(version)s/LICENSE %(installdir)s/doc',
+    'cp -r %(builddir)s/%(namelower)s-%(version)s/test_data %(installdir)s',
+    'cp -r %(builddir)s/%(namelower)s-%(version)s/config %(installdir)s',
+]
+
+sanity_check_paths = {
+    'files': ['bin/busco', 'bin/generate_plot.py'],
+    'dirs': ['test_data', 'lib/python%(pyshortver)s/site-packages/busco']
+}
+
+sanity_check_commands = [
+    "busco --help",
+    "busco -i %(installdir)s/test_data/bacteria/genome.fna --cpu %(parallel)s  --mode geno --out test_bacteria.out",
+    "busco -i %(installdir)s/test_data/eukaryota/genome.fna --cpu %(parallel)s --mode geno --out test_eukaryota.out",
+]
+
+moduleclass = 'bio'

config/easybuild/easyconfigs/b/BamTools/BamTools-2.5.2-GCC-13.2.0.eb

@@ -0,0 +1,22 @@
+# This file is an EasyBuild reciPY as per https://github.com/easybuilders/easybuild
+name = 'BamTools'
+version = '2.5.2'
+
+homepage = 'https://github.com/pezmaster31/bamtools'
+description = "BamTools provides both a programmer's API and an end-user's toolkit for handling BAM files."
+
+toolchain = {'name': 'GCC', 'version': '13.2.0'}
+toolchainopts = {'pic': True}
+
+source_urls = [GITHUB_LOWER_SOURCE]
+sources = ['v%(version)s.tar.gz']
+checksums = ['4d8b84bd07b673d0ed41031348f10ca98dd6fa6a4460f9b9668d6f1d4084dfc8']
+
+builddependencies = [
+    ('CMake', '3.27.6'),
+]
+
+# https://github.com/pezmaster31/bamtools
+github_account = 'pezmaster31'
+
+moduleclass = 'bio'

config/easybuild/easyconfigs/d/DendroPy/DendroPy-4.6.1-GCCcore-13.2.0.eb

@@ -0,0 +1,43 @@
+##
+# This file is an EasyBuild reciPY as per https://github.com/easybuilders/easybuild
+#
+# Copyright:: Copyright 2013-2014 The Cyprus Institute
+# Authors::   Thekla Loizou <[email protected]>
+# License::   MIT/GPL
+# $Id$
+#
+##
+easyblock = 'PythonPackage'
+
+name = 'DendroPy'
+version = '4.6.1'
+
+homepage = 'https://dendropy.org/'
+description = """A Python library for phylogenetics and phylogenetic computing: 
+reading, writing, simulation, processing and manipulation of phylogenetic trees 
+(phylogenies) and characters."""
+
+toolchain = {'name': 'GCCcore', 'version': '13.2.0'}
+
+sources = [SOURCE_TAR_GZ]
+checksums = ['26fcbe1cb5831301e8f1f2e15a0563620f0b8e29e6d409dd6a2a7c957dd64c16']
+
+builddependencies = [('binutils', '2.40')]
+
+dependencies = [('Python', '3.11.5')]
+
+download_dep_fail = True
+use_pip = True
+
+fix_python_shebang_for = ['bin/*.py']
+
+sanity_check_paths = {
+    'files': ['bin/sumlabels.py', 'bin/sumtrees.py'],
+    'dirs': ['lib/python%(pyshortver)s/site-packages'],
+}
+
+sanity_check_commands = ["sumtrees.py --help"]
+
+sanity_pip_check = True
+
+moduleclass = 'bio'

config/easybuild/easyconfigs/l/lpsolve/lpsolve-5.5.2.11-GCC-13.2.0.eb

@@ -0,0 +1,33 @@
+easyblock = 'CmdCp'
+
+name = 'lpsolve'
+version = '5.5.2.11'
+
+homepage = 'https://sourceforge.net/projects/lpsolve/'
+description = "Mixed Integer Linear Programming (MILP) solver"
+
+toolchain = {'name': 'GCC', 'version': '13.2.0'}
+
+source_urls = [SOURCEFORGE_SOURCE]
+sources = ['lp_solve_%(version)s_source.tar.gz']
+checksums = ['6d4abff5cc6aaa933ae8e6c17a226df0fc0b671c438f69715d41d09fe81f902f']
+
+local_lpsolve_ver = '%(version_major)s%(version_minor)s'
+start_dir = 'lpsolve%s' % local_lpsolve_ver
+
+local_comp_cmd = 'sed -i "s/^c=.*/c=\'$CC\'/g" ccc && sed -i "s/^opts=.*/opts=\'$CFLAGS\'/g" ccc && '
+local_comp_cmd += "sh ccc"
+cmds_map = [('.*', local_comp_cmd)]
+
+local_lpsolve_libname = 'liblpsolve%s' % local_lpsolve_ver
+files_to_copy = [
+    (['bin/ux64/%s.a' % local_lpsolve_libname, 'bin/ux64/%s.%s' % (local_lpsolve_libname, SHLIB_EXT)], 'lib'),
+    (['../lp*.h'], 'include'),
+]
+
+sanity_check_paths = {
+    'files': ['lib/%s.a' % local_lpsolve_libname, 'lib/%s.%s' % (local_lpsolve_libname, SHLIB_EXT)],
+    'dirs': ['include'],
+}
+
+moduleclass = 'math'