Repository navigation
Releases: 3d-omics/mt_quant
Releases · 3d-omics/mt_quant
Release list
v1.4.0
1.4.0 (2024-12-04)
Features
- add slurm, cache, test scripts (4ab571f)
- bracken (8183326)
- cache rule (e7ac7bd)
- configfile (ebac6c0)
- do bracken on all taxonomic levels, only show species in multiqc (d8efee3)
- get the aggregated counts directly from subread (7f8900e)
- use col1 and col2 to see if it reduces ram usage (2f527c2)
- use subread from wrappers (e9971d8)
Bug Fixes
- add all the bracken files to the copy in /dev/shm (c4df691)
- dont use slurm within slurm , dumbass (acf4c57)
- get the correct featurecounts file (a2beada)
- handle manually subread sample renaming; corner case in csvtk (0a15ae8)
- put clean in the preprocess group (24be3b8)
- put fastq in the correct group (e6a3462)
- put slurm executor (f4b65f6)
- put tmp folders for index and map (a3c11ab)
- remove more mentions to coverm (900e667)
- remove tmp dirs if they exist (ccf4e09)
- rerun incomplete (99941d1)
- undo bracken copy (47f72eb)
- use proper featurecounts file (d76c680)
Performance Improvements
v1.3.0
1.3.0 (2024-11-29)
Features
- coverm: mark individual files as temporary (25e4c86)
- coverm: rearrange files again (8058735)
- coverm: remove functions (015c4f8)
- htseq: do not compress output and make it temp (6d9159c)
- remove r packages, update and pin environments (5785f85)
- restructure bowtie and use csvtk to join files (acc9ddf)
- subread: introduce middle step (365d41e)
- subread: quantify trnas and rrnas too (d459a74)
- subread: solve dependency; add reports to multiqc (2b7b9a3)
- substitute custom r script for csvtk (0cacb96)
Bug Fixes
v1.2.0
1.2.0 (2024-11-12)
Features
- coverm++ (4b170ac)
- demand snakemake 8 (5cec580)
- disable crams in prerpocessing (f329758)
- disable crams in quantify (c618038)
- group jobs by sample, update profile (658ae38)
- import helpers. insert reads into prerpocess. remove per library report (af3f230)
- insert hosts into preprocess (4c285c0)
- multqc handles more files (4d44cad)
- point to base env and clean other (b1b34e0)
- refactor fastp. use wrappers (da1a88d)
- refactor quantify (e63eca6)
- refactor quantify (d59ce09)
- refactor ribodetector (cee57ef)
- refactor ribodetector (bc4f921)
- remove fastp fastqcs from report (9323aec)
- remove fastp functions (5eda525)
- separate bowtie2 env from quant (71a54a6)
- update quantify rule names, fix multiqc output (96cb3e8)
Bug Fixes
- fastp: correct group name (2751414)
- remove mention to ram (e009781)
- star: point to u1, u2 (f6344f1)
- star: point to u1, u2 .fq (fb68d67)
Performance Improvements
v1.1.2
v1.1.1
v1.1.0
v1.0.1
v1.0.0
1.0.0 (2024-06-14)
⚠ BREAKING CHANGES
- release 1.0.0
Features
- add gha updater (6dd2b43)
- add per gene counts with featureCount (3f8fd05)
- bowtie2: write index within the rule (f171f0c)
- bowtie2: write index within the rule (5f17b3f)
- compress all count data (72e9a9e)
- compute counts over genes (a222296)
- coverm: coverm + cram (dbe1b09)
- coverm: coverm + cram (869df57)
- coverm: delete dead code. -F 4 (0661d62)
- coverm: delete dead code. -F 4 (dea6a0a)
- fastp: remove process substitution and use level 9 compression (7a685dd)
- fastp: remove process substitution and use level 9 compression (a45a22f)
- fastp: remove process substitution and use level 9 compression (6834cc5)
- fastp: use process substituttion and pigz -11 (9b956cb)
- fastp: use process substituttion and pigz -11 (4f3b0b9)
- fastp: use process substituttion and pigz -11 (c49b538)
- features: add the usual suspects (a2ec8c4)
- features: move databases and mag catalogue. @ in mag mock (f9c5f54)
- get gene quantifications with htseq (1a461e0)
- ignore rstudio stuff (651d20e)
- ignore rstudio stuff (39767ac)
- params: add @ as the default mag-contig separator (e33b82d)
- params: add common coverm params (47f0caf)
- params: add common coverm params (7a00763)
- pre-commit: add conventional-pre-commit (b6df867)
- pre-commit: add R linters (b31ddec)
- pre-commit: autoupgrade (99ec679)
- pre-commit: autoupgrade (e156263)
- pre-commit: update all hooks (e7cf184)
- preprocess: set adaptors in samples. refactor. Update readme (649d019)
- quantify: refactor (be70cab)
- quantify: refactor coverm functions (6a33bc0)
- README: update repo name (d516776)
- README: update repo name (c62c2be)
- reads: refactor file and adaptor functions (ba42fec)
- reads: split into subworkflows (d6bc3ba)
- reads: split into subworkflows (1616a6f)
- report: add the definitive report rule (a708af3)
- report: raise ram (f1aaf9d)
- report: raise ram (ae130be)
- ribodetector: remove temp (86c5ac2)
- ribodetector: remove temp (0617fd3)
- star: compress within the rule, raise compression level (eca2288)
- star: compress within the rule, raise compression level (d93087b)
- star: refactor functions (6258c9b)
- star: refactor functions (85b0467)
- the big rename (0d9833f)
- update github actions (2e9d284)
- Update readme (11d0942)
- Update readme (19c9962)
- Update readme (a6e5095)
Bug Fixes
- add prefixes to htseq and subread outputs to prevent a prefix loop (974402a)
- add user and permissions to rsync (b52d0fd)
- folder: consistent folder names (89603a6)
- folder: consistent folder names (22f9066)
- folder: consistent folder names (15ba09b)
- folder: consistent folder names (446d25b)
- gha: add temporary snakemake-action (78945ad)
- kraken2: use for loop over mapfile (7a6afbf)
- kraken2: use for loop over mapfile (25faa7b)
- point to the correct branch in the badge (aa9c2b6)
- pre-commit: update hooks (b04c596)
- remove unused code (8f1b473)
- report: raise preprocess report ram (80c84bc)
- report: raise preprocess report ram (4a7f1b7)
- report: rename rules (0a2197b)
Miscellaneous Chores
- release 1.0.0 (0a18829)