Merge pull request #21919 from pavelToman/20241127164105_new_pr_MATES…

…01520241121

{bio}[foss/2023b] MATES v0.1.5-20241121, anndata v0.11.1, pybedtools v0.10.0, ...

easybuild/easyconfigs/a/anndata/anndata-0.11.1-foss-2023b.eb

@@ -0,0 +1,45 @@
+easyblock = 'PythonBundle'
+
+name = 'anndata'
+version = '0.11.1'
+
+homepage = 'https://github.com/scverse/anndata'
+description = """anndata is a Python package for handling annotated data matrices in memory and on disk,
+ positioned between pandas and xarray"""
+
+toolchain = {'name': 'foss', 'version': '2023b'}
+
+builddependencies = [
+    ('hatchling', '1.18.0'),
+]
+
+dependencies = [
+    ('Python', '3.11.5'),
+    ('Python-bundle-PyPI', '2023.10'),
+    ('SciPy-bundle', '2023.11'),
+    ('h5py', '3.11.0'),
+]
+
+use_pip = True
+sanity_pip_check = True
+
+exts_list = [
+    ('array_api_compat', '1.9.1', {
+        'checksums': ['17bab828c93c79a5bb8b867145b71fcb889686607c5672b060aef437e0359ea8'],
+    }),
+    ('natsort', '8.4.0', {
+        'checksums': ['45312c4a0e5507593da193dedd04abb1469253b601ecaf63445ad80f0a1ea581'],
+    }),
+    (name, version, {
+        'checksums': ['36bff9a85276fc5f1b9fd01f15aff9aa49408129985f42e0fca4e2c5b7fa909f'],
+    }),
+]
+
+sanity_check_paths = {
+    'files': ['bin/natsort'],
+    'dirs': ['lib/python%(pyshortver)s/site-packages'],
+}
+
+sanity_check_commands = ["natsort --help"]
+
+moduleclass = 'bio'

easybuild/easyconfigs/b/BEDTools/BEDTools-2.31.1-GCC-13.2.0.eb

@@ -0,0 +1,46 @@
+# Author: Maxime Schmitt, University of Luxembourg
+# Author: Adam Huffman, The Francis Crick Institute
+#
+# Based on the work of: Pablo Escobar Lopez
+# Swiss Institute of Bioinformatics (SIB)
+# Biozentrum - University of Basel
+
+easyblock = 'MakeCp'
+
+name = 'BEDTools'
+version = '2.31.1'
+
+homepage = 'https://bedtools.readthedocs.io/'
+description = """BEDTools: a powerful toolset for genome arithmetic.
+The BEDTools utilities allow one to address common genomics tasks such as finding feature overlaps and
+computing coverage.
+The utilities are largely based on four widely-used file formats: BED, GFF/GTF, VCF, and SAM/BAM."""
+
+toolchain = {'name': 'GCC', 'version': '13.2.0'}
+
+source_urls = ['https://github.com/arq5x/bedtools2/archive/refs/tags/']
+sources = ['v%(version)s.tar.gz']
+checksums = ['79a1ba318d309f4e74bfa74258b73ef578dccb1045e270998d7fe9da9f43a50e']
+
+builddependencies = [
+    ('Python', '3.11.5'),
+]
+dependencies = [
+    ('XZ', '5.4.4'),
+    ('zlib', '1.2.13'),
+    ('bzip2', '1.0.8'),
+    ('BamTools', '2.5.2'),
+]
+
+buildopts = 'CXX="$CXX"'
+
+files_to_copy = ['bin', 'docs', 'data', 'genomes', 'scripts', 'test']
+
+sanity_check_paths = {
+    'files': ['bin/%s' % x for x in ['bedtools', 'pairToBed', 'mergeBed', 'bedToBam', 'fastaFromBed']],
+    'dirs': files_to_copy,
+}
+
+sanity_check_commands = ['%(namelower)s --help']
+
+moduleclass = 'bio'

easybuild/easyconfigs/m/MATES/MATES-0.1.5-20241121-foss-2023b.eb

@@ -0,0 +1,53 @@
+easyblock = 'PythonBundle'
+
+name = 'MATES'
+version = '0.1.5-20241121'
+local_commit = '3846ad5'
+
+homepage = 'https://github.com/mcgilldinglab/MATES'
+description = "A Deep Learning-Based Model for Quantifying Transposable Elements in Single-Cell Sequencing Data."
+
+toolchain = {'name': 'foss', 'version': '2023b'}
+
+dependencies = [
+    ('Python', '3.11.5'),
+    ('Python-bundle-PyPI', '2023.10'),
+    ('SciPy-bundle', '2023.11'),
+    ('matplotlib', '3.8.2'),
+    ('anndata', '0.11.1'),
+    ('pybedtools', '0.10.0'),
+    ('PyTorch', '2.1.2'),
+    ('Pysam', '0.22.0'),
+    ('tqdm', '4.66.2'),
+    ('SAMtools', '1.19.2'),
+]
+
+use_pip = True
+sanity_pip_check = True
+
+exts_list = [
+    ('sorted_nearest', '0.0.39', {
+        'checksums': ['16a51d5db87ae226b47ace43c176bb672477a1b7ba8052ea9291a6356c9c69b1'],
+    }),
+    ('ncls', '0.0.68', {
+        'checksums': ['81aaa5abb123bb21797ed2f8ef921e20222db14a3ecbc61ccf447532f2b7ba93'],
+    }),
+    ('pyranges', '0.0.129', {
+        'checksums': ['bee83b4fad0062be9586668c6b0fc4270d5e761951975e018202993680071fb3'],
+    }),
+    (name, version, {
+        'modulename': 'MATES',
+        # unpin exact versions of dependencies
+        'preinstallopts': """sed -i 's/==.*//g' requirements.txt && sed -i 's/==.*/\",/g' setup.py && """,
+        'source_urls': ['https://github.com/mcgilldinglab/MATES/archive'],
+        'sources': [{'download_filename': '%s.tar.gz' % local_commit, 'filename': SOURCE_TAR_GZ}],
+        'checksums': ['40fbb87dd72ca4c9e5347f2e984f9c0a0caa817d4eee692476be71e733e76f61'],
+    }),
+]
+
+sanity_check_commands = [
+    "python -c 'from MATES import bam_processor, data_processor, MATES_model'",
+    "python -c 'from MATES import TE_quantifier, TE_quantifier_LongRead, TE_quantifier_Intronic'",
+]
+
+moduleclass = 'bio'

easybuild/easyconfigs/p/pybedtools/pybedtools-0.10.0-foss-2023b.eb

@@ -0,0 +1,25 @@
+easyblock = 'PythonPackage'
+
+name = 'pybedtools'
+version = '0.10.0'
+
+homepage = 'https://daler.github.io/pybedtools'
+description = "pybedtools wraps and extends BEDTools and offers feature-level manipulations from within Python."
+
+toolchain = {'name': 'foss', 'version': '2023b'}
+
+sources = [SOURCE_TAR_GZ]
+checksums = ['1a6fbaad23b013becc741d7d5922a2df03e391bc44ff92772ffb7dd456711161']
+
+dependencies = [
+    ('Python', '3.11.5'),
+    ('SciPy-bundle', '2023.11'),
+    ('BEDTools', '2.31.1'),
+    ('Pysam', '0.22.0'),
+]
+
+use_pip = True
+download_dep_fail = True
+sanity_pip_check = True
+
+moduleclass = 'bio'