progress on #236 NGSpeciesID

236_NGSpeciesID/Biopython-1.83-foss-2023a.eb

@@ -0,0 +1,45 @@
+##
+# Author:    Robert Mijakovic <[email protected]>
+##
+easyblock = 'PythonPackage'
+
+name = 'Biopython'
+version = '1.83'
+
+homepage = 'https://www.biopython.org'
+description = """Biopython is a set of freely available tools for biological
+ computation written in Python by an international team of developers. It is
+ a distributed collaborative effort to develop Python libraries and
+ applications which address the needs of current and future work in
+ bioinformatics. """
+
+toolchain = {'name': 'foss', 'version': '2023a'}
+
+source_urls = ['https://biopython.org/DIST']
+sources = [SOURCELOWER_TAR_GZ]
+checksums = ['78e6bfb78de63034037afd35fe77cb6e0a9e5b62706becf78a7d922b16ed83f7']
+
+dependencies = [
+    ('Python', '3.11.3'),
+    ('SciPy-bundle', '2023.07'),
+]
+
+download_dep_fail = True
+use_pip = True
+sanity_pip_check = True
+
+# Run only tests that don't require internet connection
+runtest = 'python setup.py test --offline'
+
+sanity_check_paths = {
+    'files': [],
+    'dirs': ['lib/python%(pyshortver)s/site-packages/Bio',
+             'lib/python%(pyshortver)s/site-packages/BioSQL']
+}
+
+# extra check to ensure numpy dependency is available
+sanity_check_commands = ["python -c 'import Bio.MarkovModel'"]
+
+options = {'modulename': 'Bio'}
+
+moduleclass = 'bio'

236_NGSpeciesID/Pysam-0.22.0-GCC-12.3.0.eb

@@ -0,0 +1,32 @@
+easyblock = 'PythonPackage'
+
+name = 'Pysam'
+version = '0.22.0'
+
+homepage = 'https://github.com/pysam-developers/pysam'
+description = """Pysam is a python module for reading and manipulating Samfiles.
+ It's a lightweight wrapper of the samtools C-API. Pysam also includes an interface for tabix."""
+
+toolchain = {'name': 'GCC', 'version': '12.3.0'}
+
+source_urls = ['https://github.com/pysam-developers/pysam/archive/']
+sources = ['v%(version)s.tar.gz']
+checksums = ['61b3377c5f889ddc6f6979912c3bb960d7e08407dada9cb38f13955564ea036f']
+
+builddependencies = [('Python-bundle-PyPI', '2023.06')]
+
+dependencies = [
+    ('Python', '3.11.3'),
+    ('ncurses', '6.4'),
+    ('cURL', '8.0.1'),
+    ('XZ', '5.4.2'),
+]
+
+# https://github.com/pysam-developers/pysam/issues/1146#issuecomment-1309421979
+preinstallopts = """export CFLAGS="-fPIC $CFLAGS" && """
+download_dep_fail = True
+use_pip = True
+
+sanity_pip_check = True
+
+moduleclass = 'bio'

236_NGSpeciesID/Racon-1.5.0-GCCcore-12.3.0.eb

@@ -0,0 +1,29 @@
+easyblock = 'CMakeMake'
+
+name = 'Racon'
+version = '1.5.0'
+
+homepage = 'https://github.com/lbcb-sci/racon'
+description = """Ultrafast consensus module for raw de novo genome assembly of long uncorrected reads."""
+
+toolchain = {'name': 'GCCcore', 'version': '12.3.0'}
+
+github_account = 'lbcb-sci'
+source_urls = [GITHUB_SOURCE]
+sources = ['%(version)s.tar.gz']
+checksums = ['41e362f71cc03b934f17d6e2c0d626e1b2997258261b14551586de006666424a']
+
+builddependencies = [
+    ('CMake', '3.26.3'),
+    ('binutils', '2.40'),
+    # ('git', '2.38.1', '-nodocs'),
+]
+
+sanity_check_paths = {
+    'files': ['bin/racon'],
+    'dirs': [],
+}
+
+sanity_check_commands = ['racon --help']
+
+moduleclass = 'bio'

236_NGSpeciesID/edlib-1.3.9-GCC-12.3.0.eb

@@ -0,0 +1,24 @@
+easyblock = 'PythonPackage'
+
+name = 'edlib'
+version = '1.3.9'
+
+homepage = 'https://martinsos.github.io/edlib'
+description = "Lightweight, super fast library for sequence alignment using edit (Levenshtein) distance."
+
+toolchain = {'name': 'GCC', 'version': '12.3.0'}
+
+sources = [SOURCE_TAR_GZ]
+checksums = ['64c3dfab3ebe3e759565a0cc71eb4df23cf3ce1713fd558af3c473dddc2a3766']
+
+dependencies = [
+    ('Python', '3.11.3'),
+]
+
+download_dep_fail = True
+use_pip = True
+sanity_pip_check = True
+
+moduleclass = 'bio'
+
+# TODO edlib.bycython.cpp:198:12: fatal error: longintrepr.h: No such file or directory

236_NGSpeciesID/intervaltree-python-3.1.0-GCCcore-12.3.0.eb

@@ -0,0 +1,35 @@
+# Author: Jasper Grimm (UoY)
+# Update: Petr Král (INUITS)
+
+easyblock = 'PythonPackage'
+
+name = 'intervaltree-python'
+_modname = 'intervaltree'
+version = '3.1.0'
+
+homepage = 'https://github.com/chaimleib/intervaltree'
+description = """A mutable, self-balancing interval tree. Queries may be by
+ point, by range overlap, or by range containment.
+"""
+
+toolchain = {'name': 'GCCcore', 'version': '12.3.0'}
+
+source_urls = ['https://pypi.python.org/packages/source/i/%s' % _modname]
+sources = ['%s-%s.tar.gz' % (_modname, version)]
+checksums = ['902b1b88936918f9b2a19e0e5eb7ccb430ae45cde4f39ea4b36932920d33952d']
+
+builddependencies = [('binutils', '2.40')]
+
+dependencies = [
+    ('Python', '3.11.3'),
+]
+
+use_pip = True
+download_dep_fail = True
+sanity_pip_check = True
+
+options = {'modulename': _modname}
+
+moduleclass = 'bio'
+
+# TODO intervaltree 3.1.0 requires sortedcontainers, which is not installed.

236_NGSpeciesID/medaka-1.11.3-foss-2023a.eb

@@ -0,0 +1,71 @@
+# This is a contribution from HPCNow! (http://hpcnow.com)
+# Copyright::   HPCNow!
+# Authors::     Danilo Gonzalez <[email protected]>
+# License::     GPL-v3.0
+# Updated to foss-2020b to use with artic tool
+# J. Sassmannshausen (GSTT/NHS UK)
+# Updated to 1.5.0
+# Jasper Grimm (UoY)
+# Updated: Petr Král (INUITS)
+
+easyblock = 'PythonBundle'
+
+name = 'medaka'
+version = '1.11.3'
+
+homepage = 'https://github.com/nanoporetech/medaka'
+description = "medaka is a tool to create a consensus sequence from nanopore sequencing data."
+
+toolchain = {'name': 'foss', 'version': '2023a'}
+toolchainopts = {'pic': True}
+
+builddependencies = [('Autotools', '20220317')]
+
+_minimap_ver = '2.26'
+dependencies = [
+    ('Python', '3.11.3'),
+    ('Python-bundle-PyPI', '2023.06'),  # includes cffi
+    # tensorflow~=2.10.0 required by medaka 1.9.1, see requirements.txt
+    ('TensorFlow', '2.13.0'),
+    ('Pysam', '0.21.0'),  # TODO
+    ('SAMtools', '1.18'),
+    ('minimap2', _minimap_ver),
+    ('HTSlib', '1.18'),  # for tabix, bgzip
+    ('Racon', '1.5.0'),  # TODO
+    ('edlib', '1.3.9'),  # TODO
+    ('pyspoa', '0.2.1'),
+    ('python-parasail', '1.3.4'),  # TODO
+    ('ont-fast5-api', '4.1.1'),  # TODO
+    ('WhatsHap', '2.1'),  # TODO
+    ('intervaltree-python', '3.1.0'),  # TODO
+    ('BCFtools', '1.18'),
+]
+
+use_pip = True
+sanity_pip_check = True
+
+exts_list = [
+    ('mappy', _minimap_ver, {
+        'checksums': ['e53fbe9a3ea8762a64b8103f4f779c9fb16d418eaa0a731f45cebc83867a9b71'],
+    }),
+    ('wurlitzer', '3.0.3', {
+        'checksums': ['224f5fe70618be3872c05dfddc8c457191ec1870654596279fcc1edadebe3e5b'],
+    }),
+    (name, version, {
+        'checksums': ['4440762a17ddd66806ddbd7c3218140caa234b96a8c919ed54d7243d3e4a5dd1'],
+        # Some requirements are too strict.
+        'preinstallopts': "sed -i 's/tensorflow.*/tensorflow/g;s/cffi==/cffi>=/g' requirements.txt && ",
+    }),
+]
+
+sanity_check_paths = {
+    'files': ['bin/medaka', 'bin/medaka_consensus', 'bin/medaka_version_report'],
+    'dirs': ['lib/python%(pyshortver)s/site-packages'],
+}
+
+sanity_check_commands = [
+    "medaka --help",
+    "medaka_version_report",
+]
+
+moduleclass = 'bio'

236_NGSpeciesID/minimap2-2.26-GCCcore-12.3.0.eb

@@ -0,0 +1,54 @@
+# This file is an EasyBuild reciPY as per https://github.com/easybuilders/easybuild
+# Adam Huffman
+# DeepThought, Flinders University
+# Updated to 2.22
+# R.QIAO
+
+# Update Petr Král (INUITS)
+easyblock = 'MakeCp'
+
+name = 'minimap2'
+version = '2.26'
+
+homepage = 'https://github.com/lh3/minimap2'
+description = """Minimap2 is a fast sequence mapping and alignment
+program that can find overlaps between long noisy reads, or map long
+reads or their assemblies to a reference genome optionally with detailed
+alignment (i.e. CIGAR). At present, it works efficiently with query
+sequences from a few kilobases to ~100 megabases in length at an error
+rate ~15%. Minimap2 outputs in the PAF or the SAM format. On limited
+test data sets, minimap2 is over 20 times faster than most other
+long-read aligners. It will replace BWA-MEM for long reads and contig
+alignment."""
+
+toolchain = {'name': 'GCCcore', 'version': '12.3.0'}
+
+source_urls = ['https://github.com/lh3/%(name)s/releases/download/v%(version)s/']
+sources = ['%(name)s-%(version)s.tar.bz2']
+checksums = ['6a588efbd273bff4f4808d5190957c50272833d2daeb4407ccf4c1b78143624c']
+
+builddependencies = [('binutils', '2.40')]
+
+dependencies = [('zlib', '1.2.13')]
+
+buildopts = 'CC="${CC}" CFLAGS="${CFLAGS}" INCLUDES="${CPPFLAGS}"'
+
+files_to_copy = [
+    (['%(name)s'], 'bin'),
+    (['lib%(name)s.a'], 'lib'),
+    (['*.h'], 'include'),
+    'LICENSE.txt', 'NEWS.md', 'README.md',
+    (['%(name)s.1'], 'share/man/man1')
+]
+
+sanity_check_paths = {
+    'files': ['bin/%(name)s', 'lib/lib%(name)s.a'],
+    'dirs': ['include']
+}
+
+sanity_check_commands = [
+    "minimap2 --help",
+    "cd %(builddir)s/minimap2-%(version)s && minimap2 -a test/MT-human.fa test/MT-orang.fa > test.sam",
+]
+
+moduleclass = 'bio'

236_NGSpeciesID/ont-fast5-api-4.1.2-foss-2023a.eb

@@ -0,0 +1,43 @@
+easyblock = 'PythonBundle'
+
+name = 'ont-fast5-api'
+version = '4.1.2'
+
+homepage = 'https://github.com/nanoporetech/ont_fast5_api'
+description = "ont_fast5_api is a simple interface to HDF5 files of the Oxford Nanopore .fast5 file format."
+
+toolchain = {'name': 'foss', 'version': '2023a'}
+
+dependencies = [
+    ('Python', '3.11.3'),
+    ('SciPy-bundle', '2023.07'),
+    ('h5py', '3.9.0'),
+]
+
+use_pip = True
+
+exts_list = [
+    ('progressbar33', '2.4', {
+        'modulename': 'progressbar',
+        'checksums': ['51fe0d9b3b4023db2f983eeccdfc8c9846b84db8443b9bee002c7f58f4376eff'],
+    }),
+    (name, version, {
+        'checksums': ['c7c59c6100e992ef8bc239cdf91f7a8ab46abf57ecd689f94b2b98e72a9e9472'],
+    }),
+]
+
+sanity_check_paths = {
+    'files': ['bin/compress_fast5', 'bin/fast5_subset', 'bin/multi_to_single_fast5', 'bin/single_to_multi_fast5'],
+    'dirs': [''],
+}
+
+sanity_check_commands = [
+    "compress_fast5 --help",
+    "fast5_subset --help",
+    "multi_to_single_fast5 --help",
+    "single_to_multi_fast5 --help",
+]
+
+sanity_pip_check = True
+
+moduleclass = 'bio'

236_NGSpeciesID/parasail-2.6.2-GCC-12.3.0.eb

@@ -0,0 +1,27 @@
+easyblock = 'CMakeMake'
+
+name = 'parasail'
+version = '2.6.2'
+
+homepage = 'https://github.com/jeffdaily/parasail'
+description = """parasail is a SIMD C (C99) library containing implementations
+ of the Smith-Waterman (local), Needleman-Wunsch (global), and semi-global
+ pairwise sequence alignment algorithms. """
+
+toolchain = {'name': 'GCC', 'version': '12.3.0'}
+
+# https://github.com/jeffdaily/parasail
+github_account = 'jeffdaily'
+source_urls = [GITHUB_SOURCE]
+sources = ['v%(version)s.tar.gz']
+checksums = ['9057041db8e1cde76678f649420b85054650414e5de9ea84ee268756c7ea4b4b']
+
+builddependencies = [('CMake', '3.26.3')]
+
+sanity_check_paths = {
+    'files': ['bin/parasail_aligner', 'bin/parasail_stats',
+              'lib/libparasail.%s' % SHLIB_EXT, 'include/parasail.h'],
+    'dirs': [],
+}
+
+moduleclass = 'bio'