Merge pull request #19779 from pavelToman/20240202110825_new_pr_phylu…

…ce173

{bio}[foss/2023a] phyluce v1.7.3

easybuild/easyconfigs/a/ABySS/ABySS-2.3.7-foss-2023a.eb

@@ -0,0 +1,34 @@
+easyblock = 'ConfigureMake'
+
+name = 'ABySS'
+version = '2.3.7'
+
+homepage = 'https://www.bcgsc.ca/platform/bioinfo/software/abyss'
+description = """Assembly By Short Sequences - a de novo, parallel, paired-end sequence assembler"""
+
+toolchain = {'name': 'foss', 'version': '2023a'}
+toolchainopts = {'usempi': True, 'cstd': 'c++17'}
+
+source_urls = ['http://github.com/bcgsc/abyss/releases/download/%(version)s/']
+sources = [SOURCELOWER_TAR_GZ]
+checksums = ['ba37780e79ec3aa359b6003e383caef13479a87f4d0022af01b86398f9ffca1f']
+
+dependencies = [
+    ('Autoconf', '2.71'),
+    ('Automake', '1.16.5'),
+    ('Boost', '1.82.0'),
+    ('sparsehash', '2.0.4'),
+    ('btllib', '1.7.0'),
+]
+
+preconfigopts = "./autogen.sh && "
+configopts = 'CXXFLAGS="$CXXFLAGS -Wno-error"'
+
+sanity_check_paths = {
+    'files': ["bin/ABYSS", "bin/ABYSS-P"],
+    'dirs': []
+}
+
+sanity_check_commands = ['ABYSS --help']
+
+moduleclass = 'bio'

easybuild/easyconfigs/b/btllib/btllib-1.7.0-GCC-12.3.0.eb

@@ -0,0 +1,43 @@
+easyblock = 'CmdCp'
+
+name = 'btllib'
+version = '1.7.0'
+
+homepage = 'https://github.com/bcgsc/btllib'
+description = """Bioinformatics Technology Lab common code library"""
+
+toolchain = {'name': 'GCC', 'version': '12.3.0'}
+toolchainopts = {'openmp': True}
+
+sources = [{
+    'filename': '%(name)s-%(version)s.tar.gz',
+    'git_config': {
+        'url': 'https://github.com/bcgsc',
+        'repo_name': 'btllib',
+        'tag': 'v%(version)s',
+        'recursive': True,
+        'keep_git_dir': True,
+    }
+}]
+checksums = [None]
+
+dependencies = [
+    ('Python', '3.11.3'),
+    ('Meson', '1.1.1'),
+    ('Ninja', '1.11.1'),
+    ('CMake', '3.26.3'),
+    ('SAMtools', '1.18'),
+]
+
+cmds_map = [('.*', "./compile")]
+
+files_to_copy = [(['install/bin/*'], 'bin'), (['install/lib/*'], 'lib'), (['install/include/*'], 'include')]
+
+sanity_check_paths = {
+    'files': ['bin/randseq', 'bin/indexlr', 'bin/mi_bf_generate'],
+    'dirs': [],
+}
+
+sanity_check_commands = ['randseq --help', 'indexlr --help']
+
+moduleclass = 'lang'

easybuild/easyconfigs/l/LASTZ/LASTZ-1.04.22-GCC-12.3.0.eb

@@ -0,0 +1,29 @@
+easyblock = 'ConfigureMake'
+
+name = 'LASTZ'
+version = '1.04.22'
+
+homepage = 'https://github.com/lastz/lastz'
+description = """ LASTZ is a program for aligning DNA sequences, a pairwise aligner. Originally designed to handle
+ sequences the size of human chromosomes and from different species, it is also useful for sequences produced by NGS 
+ sequencing technologies such as Roche 454.
+"""
+
+toolchain = {'name': 'GCC', 'version': '12.3.0'}
+
+source_urls = [GITHUB_LOWER_SOURCE]
+sources = ['%(version)s.tar.gz']
+checksums = ['4c829603ba4aed7ddf64255b528cd88850e4557382fca29580d3576c25c5054a']
+
+skipsteps = ['configure']
+
+buildopts = "allowBackToBackGaps=ON"
+
+installopts = 'installDir=%(installdir)s/bin'
+
+sanity_check_paths = {
+    'files': ['bin/lastz', 'bin/lastz_D'],
+    'dirs': [],
+}
+
+moduleclass = 'bio'

easybuild/easyconfigs/p/Pilon/Pilon-1.23-Java-11.eb

@@ -10,21 +10,29 @@ description = " Pilon is an automated genome assembly improvement and variant de
 toolchain = SYSTEM
 
 source_urls = ['https://github.com/broadinstitute/pilon/releases/download/v%(version)s/']
-sources = ['%(namelower)s-%(version)s.jar']
+sources = ['pilon-%(version)s.jar']
 checksums = ['bde1d3c8da5537abbc80627f0b2a4165c2b68551690e5733a6adf62413b87185']
 
 dependencies = [('Java', '11')]
 
-sanity_check_paths = {
-    'files': ['pilon-%(version)s.jar'],
-    'dirs': [],
-}
-
-postinstallcmds = ['cd %(installdir)s && ln -s %(namelower)s-%(version)s.jar pilon.jar']
+postinstallcmds = [
+    "cd %(installdir)s && ln -s pilon-%(version)s.jar pilon.jar",
+    "mkdir -p %(installdir)s/bin",
+    "echo '#!/bin/sh' > %(installdir)s/bin/pilon",
+    "echo 'java -jar -Xmx8G %(installdir)s/pilon-%(version)s.jar $@' >> %(installdir)s/bin/pilon",
+    "chmod a+rx %(installdir)s/bin/pilon"
+]
 
 modloadmsg = """
 To execute Pilon run: java -Xmx8G -jar $EBROOTPILON/pilon.jar
 Adjust the memory value according to the size of your input files
 """
 
+sanity_check_paths = {
+    'files': ['bin/pilon', 'pilon.jar'],
+    'dirs': [],
+}
+
+sanity_check_commands = ["pilon --help"]
+
 moduleclass = 'bio'

easybuild/easyconfigs/p/phyluce/phyluce-1.7.3-foss-2023a.eb

@@ -0,0 +1,138 @@
+easyblock = 'PythonBundle'
+
+name = 'phyluce'
+version = '1.7.3'
+
+homepage = 'https://github.com/faircloth-lab/phyluce'
+description = """
+phyluce is a software package for working with data generated from sequence capture of UCE 
+(ultra-conserved element) loci, as first published in [BCF2012]. Specifically, phyluce is a suite of programs to:
+1) assemble raw sequence reads from Illumina platforms into contigs
+2) determine which contigs represent UCE loci
+3) filter potentially paralagous UCE loci
+4) generate different sets of UCE loci across taxa of interest
+"""
+
+toolchain = {'name': 'foss', 'version': '2023a'}
+
+dependencies = [
+    ('Python', '3.11.3'),
+    ('Biopython', '1.83'),
+    ('Python-bundle-PyPI', '2023.06'),
+    ('snakemake', '8.4.2'),
+    ('DendroPy', '4.6.1'),
+    ('bx-python', '0.10.0'),
+    ('ABySS', '2.3.7'),
+    ('BCFtools', '1.18'),
+    ('BEDTools', '2.31.0'),
+    ('BWA', '0.7.17'),
+    ('Gblocks', '0.91b', '', SYSTEM),
+    ('LASTZ', '1.04.22'),
+    ('MAFFT', '7.520', '-with-extensions'),
+    ('MUSCLE', '5.1.0'),
+    ('Pilon', '1.23', '-Java-11', SYSTEM),
+    ('RAxML-NG', '1.2.0'),
+    ('SAMtools', '1.18'),
+    ('seqtk', '1.4'),
+    ('SPAdes', '3.15.4'),
+    ('trimAl', '1.4.1'),
+    ('Velvet', '1.2.10', '-mt-kmer_191')
+]
+
+exts_list = [
+    (name, version, {
+        'source_urls': ['https://github.com/faircloth-lab/phyluce/archive/refs/tags/'],
+        'sources': ['v%(version)s.tar.gz'],
+        'patches': [
+            '%(name)s-%(version)s_fix-config-paths.patch',
+            '%(name)s-%(version)s_removal-Bio-Alphabet.patch',
+        ],
+        'checksums': [
+            {'v1.7.3.tar.gz': '931fd512730bb101266b27728576df00277858d256e1ab30d64b474588362e11'},
+            {'phyluce-1.7.3_fix-config-paths.patch':
+             '22d5467d7498a9b2b718660e9580fb41986964e122e4c3baf851a324a968ae5d'},
+            {'phyluce-1.7.3_removal-Bio-Alphabet.patch':
+             '9401f8b8052924476d2a73d8ac2672e750eadf11618c672d9679f3c99234049d'},
+        ],
+    }),
+]
+
+use_pip = True
+sanity_pip_check = True
+
+sanity_check_paths = {
+    'files': [],
+    'dirs': ['lib/python%(pyshortver)s/site-packages'],
+}
+
+sanity_check_commands = ["%s --help" % x for x in [
+    "phyluce_align_extract_taxa_from_alignments",
+    "phyluce_align_extract_taxon_fasta_from_alignments",
+    "phyluce_align_filter_alignments",
+    "phyluce_align_format_concatenated_phylip_for_paml",
+    "phyluce_align_get_align_summary_data",
+    "phyluce_align_get_gblocks_trimmed_alignments_from_untrimmed",
+    "phyluce_align_get_incomplete_matrix_estimates",
+    "phyluce_align_get_informative_sites",
+    "phyluce_align_get_only_loci_with_min_taxa",
+    "phyluce_align_get_ry_recoded_alignments",
+    "phyluce_align_get_smilogram_from_alignments",
+    "phyluce_align_get_taxon_locus_counts_in_alignments",
+    "phyluce_align_get_trimal_trimmed_alignments_from_untrimmed",
+    "phyluce_align_move_align_by_conf_file",
+    "phyluce_align_randomly_sample_and_concatenate",
+    "phyluce_align_reduce_alignments_with_raxml",
+    "phyluce_align_remove_empty_taxa",
+    "phyluce_align_remove_locus_name_from_files",
+    "phyluce_align_screen_alignments_for_problems",
+    "phyluce_align_seqcap_align",
+    "phyluce_align_split_concat_nexus_to_loci",
+    "phyluce_assembly_assemblo_abyss",
+    "phyluce_assembly_assemblo_spades",
+    "phyluce_assembly_assemblo_velvet",
+    "phyluce_assembly_explode_get_fastas_file",
+    "phyluce_assembly_extract_contigs_to_barcodes",
+    "phyluce_assembly_get_bed_from_lastz",
+    "phyluce_assembly_get_fasta_lengths",
+    "phyluce_assembly_get_fastas_from_match_counts",
+    "phyluce_assembly_get_fastq_lengths",
+    "phyluce_assembly_get_match_counts",
+    "phyluce_assembly_match_contigs_to_barcodes",
+    "phyluce_genetrees_get_mean_bootrep_support",
+    "phyluce_genetrees_generate_multilocus_bootstrap_count",
+    "phyluce_genetrees_rename_tree_leaves",
+    "phyluce_genetrees_sort_multilocus_bootstraps",
+    "phyluce_genetrees_get_tree_counts",
+    "phyluce_ncbi_chunk_fasta_for_ncbi",
+    "phyluce_ncbi_prep_uce_align_files_for_ncbi",
+    "phyluce_probe_easy_lastz",
+    "phyluce_probe_get_genome_sequences_from_bed",
+    "phyluce_probe_get_locus_bed_from_lastz_files",
+    "phyluce_probe_get_multi_fasta_table",
+    "phyluce_probe_get_multi_merge_table",
+    "phyluce_probe_get_probe_bed_from_lastz_files",
+    "phyluce_probe_get_screened_loci_by_proximity",
+    "phyluce_probe_get_subsets_of_tiled_probes",
+    "phyluce_probe_get_tiled_probe_from_multiple_inputs",
+    "phyluce_probe_get_tiled_probes",
+    "phyluce_probe_query_multi_fasta_table",
+    "phyluce_probe_query_multi_merge_table",
+    "phyluce_probe_reconstruct_uce_from_probe",
+    "phyluce_probe_remove_duplicate_hits_from_probes_using_lastz",
+    "phyluce_probe_remove_overlapping_probes_given_config",
+    "phyluce_probe_run_multiple_lastzs_sqlite",
+    "phyluce_probe_slice_sequence_from_genomes",
+    "phyluce_probe_strip_masked_loci_from_set",
+    "phyluce_utilities_combine_reads",
+    "phyluce_utilities_filter_bed_by_fasta",
+    "phyluce_utilities_get_bed_from_fasta",
+    "phyluce_utilities_merge_multiple_gzip_files",
+    "phyluce_utilities_merge_next_seq_gzip_files",
+    "phyluce_utilities_replace_many_links",
+    "phyluce_utilities_unmix_fasta_reads",
+    "phyluce_workflow",
+]]
+
+modextrapaths = {'PYTHONPATH': 'lib/python%(pyshortver)s/site-packages/phyluce'}
+
+moduleclass = 'bio'

easybuild/easyconfigs/p/phyluce/phyluce-1.7.3_fix-config-paths.patch

@@ -0,0 +1,53 @@
+fix path of binaries in config file
+author: Pavel Tomanek (INUITS)
+--- config/phyluce.conf.orig	2024-02-01 11:38:15.279943701 +0100
++++ config/phyluce.conf	2024-02-01 14:52:31.565662528 +0100
+@@ -1,27 +1,27 @@
+ [binaries]
+-abyss:$CONDA/bin/ABYSS
+-abyss-pe:$CONDA/bin/abyss-pe
+-bcftools:$CONDA/bin/bcftools
+-bedtools:$CONDA/bin/bedtools
+-bwa:$CONDA/bin/bwa
+-gblocks:$CONDA/bin/Gblocks
+-lastz:$CONDA/bin/lastz
+-mafft:$CONDA/bin/mafft
+-muscle:$CONDA/bin/muscle
+-pilon:$CONDA/bin/pilon
+-raxml-ng:$CONDA/bin/raxml-ng
+-samtools:$CONDA/bin/samtools
+-seqtk:$CONDA/bin/seqtk
+-spades:$CONDA/bin/spades.py
+-trimal:$CONDA/bin/trimal
+-velvetg:$CONDA/bin/velvetg
+-velveth:$CONDA/bin/velveth
+-snakemake:$CONDA/bin/Snakemake
++abyss:$EBROOTABYSS/bin/ABYSS
++abyss-pe:$EBROOTABYSS/bin/abyss-pe
++bcftools:$EBROOTBCFTOOLS/bin/bcftools
++bedtools:$EBROOTBEDTOOLS/bin/bedtools
++bwa:$EBROOTBWA/bin/bwa
++gblocks:$EBROOTGBLOCKS/Gblocks
++lastz:$EBROOTLASTZ/bin/lastz
++mafft:$EBROOTMAFFT/bin/mafft
++muscle:$EBROOTMUSCLE/bin/muscle
++pilon:$EBROOTPILON/bin/pilon
++raxml-ng:$EBROOTRAXMLMINNG/bin/raxml-ng
++samtools:$EBROOTSAMTOOLS/bin/samtools
++seqtk:$EBROOTSEQTK/bin/seqtk
++spades:$EBROOTSPADES/bin/spades.py
++trimal:$EBROOTTRIMAL/bin/trimal
++velvetg:$EBROOTVELVET/bin/velvetg
++velveth:$EBROOTVELVET/bin/velveth
++snakemake:$EBROOTSNAKEMAKE/bin/snakemake
+
+ [workflows]
+-mapping:$WORKFLOWS/mapping/Snakefile
+-correction:$WORKFLOWS/contig-correction/Snakefile
+-phasing:$WORKFLOWS/phasing/Snakefile
++mapping:$EBROOTPHYLUCE/phyluce/workflows/mapping/Snakefile
++correction:$EBROOTPHYLUCE/phyluce/workflows/contig-correction/Snakefile
++phasing:$EBROOTPHYLUCE/phyluce/workflows/phasing/Snakefile
+
+ #----------------
+ #    Advanced

easybuild/easyconfigs/p/phyluce/phyluce-1.7.3_removal-Bio-Alphabet.patch

@@ -0,0 +1,22 @@
+removal of the Bio.Alphabet - it is not part of Biopython anymore (since v1.78)
+see also https://biopython.org/wiki/Alphabet
+author: Pavel Tomanek (INUITS)
+--- bin/ncbi/phyluce_ncbi_prep_uce_align_files_for_ncbi.orig	2024-02-01 15:53:37.732931000 +0100
++++ bin/ncbi/phyluce_ncbi_prep_uce_align_files_for_ncbi	2024-02-01 16:16:42.885167000 +0100
+@@ -18,7 +18,6 @@
+ from Bio import AlignIO
+ from Bio.Seq import Seq
+ from Bio.SeqRecord import SeqRecord
+-from Bio.Alphabet import IUPAC
+
+ from phyluce import ncbi
+ from phyluce.log import setup_logging
+@@ -127,7 +126,7 @@
+                             .upper()
+                         )
+                         new_record = SeqRecord(
+-                            Seq(new_seq, IUPAC.IUPACAmbiguousDNA()),
++                            Seq(new_seq),
+                             id=new_id,
+                             name="",
+                             description="",