ccsmeth
========
Changelog
-------------
v0.5.0
----------
bug fixes
change feature QUAL to SNratio, didn't work though?
hid '--n_vocab' and '--n_embed' in model params
add attbigru2s2/attbilstm2s2 models
update new version of pretrained model params
v0.4.1
----------
handle ValueError when fetching bam items in call_freqb module
change pytorch max version to 1.12.1
using logging instead of print/sys.stderr.write
v0.4.0
----------
optimieze call_mods module, faster generating of modbam file
v0.3.4
----------
replace 'numpy.float' by 'float" in extract_features and call_mods modules
v0.3.3
----------
more robust operation for MM/ML tags
update the call_mods model (v1 to v2), use shared params for both rnn and attention
update aggregate model (v2 to v2p)
v0.3.2
----------
fix bug (0-pos CG in reverse strand) of call_freqb module
add output check at the start of call_mods/call_freqb
update aggregate model
v0.3.1
----------
optimize call_mods module
add aggregate mode in call_freqb module
release a stable call_mods model and a stable call_freqb aggregate model
v0.3.0
----------
add multi-threads support for reading .bam by pysam
force mod prob discrete integer in [0, 255], which means mod prob in [0, 1), in ML tag
fix align_hifi module, enable minimap2 and bwa
update requirements
v0.2.3
----------
more options in train module
multi-gpu support in call_mods and trainm modules
add denoise module
Note: skip 0.2.2, cause 0.2.2 has been deleted in pypi (https://pypi.org/manage/project/ccsmeth/history/)
v0.2.1
----------
minor fixes
change default mode in extract module
improvements in call_freqb module
v0.2.0
----------
use hifi reads instead of subreads as input
modbam support
use (chrom, pos, strand) as key instead of (chrom, pos) to handle CG mismatch in CCS read when call_freq
v0.1.2
----------
add pbmm2 as default aligner
enable gzip output
v0.1.1
----------
bug fixes
v0.1.0
----------
initialize
Raw data
{
"_id": null,
"home_page": "https://github.com/PengNi/ccsmeth",
"name": "ccsmeth",
"maintainer": "",
"docs_url": null,
"requires_python": "",
"maintainer_email": "",
"keywords": "methylation,pacbio,neural network",
"author": "Peng Ni",
"author_email": "543943952@qq.com",
"download_url": "https://files.pythonhosted.org/packages/1c/d9/d0696bb296d3dabee73d0284d1490635e8ba79ddbe4528970544a6fb9367/ccsmeth-0.5.0.tar.gz",
"platform": "Linux",
"description": "ccsmeth\r\n========\r\n\r\n\r\nChangelog\r\n-------------\r\nv0.5.0\r\n----------\r\nbug fixes\r\n\r\nchange feature QUAL to SNratio, didn't work though?\r\n\r\nhid '--n_vocab' and '--n_embed' in model params\r\n\r\nadd attbigru2s2/attbilstm2s2 models\r\n\r\nupdate new version of pretrained model params\r\n\r\n\r\nv0.4.1\r\n----------\r\nhandle ValueError when fetching bam items in call_freqb module\r\n\r\nchange pytorch max version to 1.12.1\r\n\r\nusing logging instead of print/sys.stderr.write\r\n\r\n\r\nv0.4.0\r\n----------\r\noptimieze call_mods module, faster generating of modbam file\r\n\r\n\r\nv0.3.4\r\n----------\r\nreplace 'numpy.float' by 'float\" in extract_features and call_mods modules\r\n\r\n\r\nv0.3.3\r\n----------\r\nmore robust operation for MM/ML tags\r\n\r\nupdate the call_mods model (v1 to v2), use shared params for both rnn and attention\r\n\r\nupdate aggregate model (v2 to v2p)\r\n\r\n\r\nv0.3.2\r\n----------\r\nfix bug (0-pos CG in reverse strand) of call_freqb module\r\n\r\nadd output check at the start of call_mods/call_freqb\r\n\r\nupdate aggregate model\r\n\r\n\r\nv0.3.1\r\n----------\r\noptimize call_mods module\r\n\r\nadd aggregate mode in call_freqb module\r\n\r\nrelease a stable call_mods model and a stable call_freqb aggregate model\r\n\r\n\r\nv0.3.0\r\n----------\r\nadd multi-threads support for reading .bam by pysam\r\n\r\nforce mod prob discrete integer in [0, 255], which means mod prob in [0, 1), in ML tag\r\n\r\nfix align_hifi module, enable minimap2 and bwa\r\n\r\nupdate requirements\r\n\r\n\r\nv0.2.3\r\n----------\r\nmore options in train module\r\n\r\nmulti-gpu support in call_mods and trainm modules\r\n\r\nadd denoise module\r\n\r\nNote: skip 0.2.2, cause 0.2.2 has been deleted in pypi (https://pypi.org/manage/project/ccsmeth/history/)\r\n\r\n\r\nv0.2.1\r\n----------\r\nminor fixes\r\n\r\nchange default mode in extract module\r\n\r\nimprovements in call_freqb module\r\n\r\n\r\nv0.2.0\r\n----------\r\nuse hifi reads instead of subreads as input\r\n\r\nmodbam support\r\n\r\nuse (chrom, pos, strand) as key instead of (chrom, pos) to handle CG mismatch in CCS read when call_freq\r\n\r\n\r\nv0.1.2\r\n----------\r\nadd pbmm2 as default aligner\r\n\r\nenable gzip output\r\n\r\n\r\nv0.1.1\r\n----------\r\nbug fixes\r\n\r\n\r\nv0.1.0\r\n----------\r\ninitialize\r\n",
"bugtrack_url": null,
"license": "BSD-3-Clause-Clear license",
"summary": "Detecting DNA methylation from PacBio CCS reads",
"version": "0.5.0",
"project_urls": {
"Download": "https://github.com/PengNi/ccsmeth/archive/refs/tags/0.5.0.tar.gz",
"Homepage": "https://github.com/PengNi/ccsmeth"
},
"split_keywords": [
"methylation",
"pacbio",
"neural network"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "1cd9d0696bb296d3dabee73d0284d1490635e8ba79ddbe4528970544a6fb9367",
"md5": "59bc63530646c40e74246f2d69a550fd",
"sha256": "58737e28070ab4c18fb8594217d490e4d41183b2ced5e5328162f6d688e6ad9e"
},
"downloads": -1,
"filename": "ccsmeth-0.5.0.tar.gz",
"has_sig": false,
"md5_digest": "59bc63530646c40e74246f2d69a550fd",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 81544,
"upload_time": "2024-01-30T12:19:03",
"upload_time_iso_8601": "2024-01-30T12:19:03.236984Z",
"url": "https://files.pythonhosted.org/packages/1c/d9/d0696bb296d3dabee73d0284d1490635e8ba79ddbe4528970544a6fb9367/ccsmeth-0.5.0.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-01-30 12:19:03",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "PengNi",
"github_project": "ccsmeth",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"requirements": [],
"lcname": "ccsmeth"
}
JSON
