All notable changes to Rp-Bp will be documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
- Pandas
str.split
API insummarize_rpbp_predictions.py
- Requires pbiotools >= 4.0.2
riboseq_sample_name_map
,riboseq_condition_name_map
torpbp_predictions_dashboard.py
- Exception for duplicated transcript ids with de novo annotation.
- STAR output
- Redundant transcripts with de novo annotation in
summarize_rpbp_predictions.py
- ORF numbers in labels
- I/O: ensure bool in config, remove symlinks
- Documentation
- Missing GTF attributes
- Dash assets to MANIFEST
- Upgrade all dependencies/required packages, install via bioconda
- Location of (compiled) Stan models
- PyStan 2 to CmdStanPy API
- Handling of ORF labels
- Documentation
- PyStan dependency
- Fitted parameters from output
_location
and_scale
orf_type
filtering for translation prediction- GFF3 support
- Unfiltered predictions, unless [--write-unfiltered] is set
- QTI-seq, various plotting/reporting scripts, notebooks
- Stan models transformed data/parameters name/definition (no-periodicity.stan, gaussian-naive-bayes.stan)
- CmdStanPy logging
- Regression tests, pre-commit, GitHub CI workflow, etc.
- Options handling for
prepare_rpbp_genome
- Retrieve gene_id, etc. from annotations
- Dashboad applications to replace PDF reports
- Data wrangling:
summarize_rpbp_profile_construction.py
, andsummarize_rpbp_predictions.py
- ribo_utils
- orf_profile_construction scripts from pbiotools.ribo
- Configure setup() using setup.cfg, add
pbio
, update install via requirements.
- Config file for RTD
- Module with all defaults parameters, handling options for external dependencies
- Add option to force the pickling of the Stan models during installation ISSUE #88.
- Build RTD documentation (add
autodoc_mock_imports
, emptydocs/requirements.txt
) - Missing parameters in function calls, hard coded defaults
- YAMLLoadWarning
- Fall back to multiprocessing in
estimate_orf_bayes_factors
, andextract_orf_profiles
- Stan assignment operator in models
- ORF labels in
prepare_rpbp_genome
,label_orfs
andextract_orf_coordinates
, see ISSUE #96.
- Transcript exons file (unused)
- Example dataset updated and moved to new address. Updated documentation.
predict_translated_orfs
andselect_final_prediction_set
modified such that Rp-chi pipeline is used only ifchi_square_only
is in config file, otherwise Rp-Bp is used.- Handling of options passed to flexbar via arguments to
run_all_rpbp_instances
,run_rpbp_pipeline
,create_orf_profiles
andcreate_base_genome_profile
. - Handling of additional options passed to star via arguments, mainly
to
create_base_genome_profile
.
- Removed config option
seqids_to_keep
for compatibility withextract_metagene_profiles
that does not use this option anymore. This has also been removed from the example. Updated documentation. - Removed fields
delta_l
anddelta_h
from output inestimate_orf_bayes_factors
. - Removed deprecated function calls from
analysis/profile_construction/get_all_read_filtering_counts
,prepare_rpbp_genome
,create_base_genome_profile
- Utility for supressing pystan (or other compiled function) output. For more details see Issue #10.
- Added limited support for GFF3 specifications, in particular this requires removing the
STOP codons from the CDSs and passing the right option to
STAR
.Prepare_rpbp_genome
andcreate_base_genome_profile
modified. - If using de_novo, all annotations are concatenated automatically in preparation of running the main pipeline.
- Minor changes to plotting options, typos and/or redundant features in
rpbp/analysis/profile_construction/create_rpbp_preprocessing_report
andrpbp/analysis/profile_construction/visualize_metagene_profile_bayes_factor
. See ISSUE #87. In progress. - Added default sjdbOverhang option to star during mapping.
- Removed deprecated function calls from
analysis/rpbp_predictions/add_mygene_info_to_orfs
,analysis/find_differential_micropeptides
,analysis/proteomics/get_orf_peptide_matches
andanalysis/proteomics/filter_nonunique_peptide_matches
- In
create_rpbp_preprocessing_report.create_figures
call tofilenames.get_riboseq_profiles
modified to fix file name reference issue. - In
create_rpbp_preprocessing_report
some figures were referenced before they were created, as a temporary fix, the functioncreate_figures
is now called earlier. - Minor typos in some of the analysis scripts.
- Flag added to differentiate between exons file (list of exons for any given transcript), and orfs file (technically, a list of exons per orf). See Issue #59 for more details.
- Dependencies to new repo locations.
- Reference to helper script to install prerequisite programs. See Issue #75 for more details.
-
Dependencies to new
pymisc-utils
andpybio-utils
repos. See Issue #74 for more details. -
Version specification of prereqs
cluster-subcodon-counts
. See Issue #78 for more details.
- Documentation for using custom alignment files. Please see Issue #73 for more details.
- Documentation links to old ipython notebooks. Please see Issue #71 for more details.
-
Incorrect read filtering report. See Issue #69 for more details.
-
Broken download link for example.
- Incorrect alignment of ORF data frame and profile matrix. See Issue #54 for more details.
- ORF coordinates for start codon on exon boundary. See Issue #64 after the reopening for more details.
- Predictions report formatting
- Guessing read filtering plot tick marks. See Issue #60 for details.
- ORF coordinates when start and stop codons fall on an exon boundary. See Issue #64 for more details.
- Documentation for the analysis scripts
- Documentation for creating reference indices
- Checks that annotation files exist before beginning index creation. See Issue #62
- Remove unnecessary
--tmp
argument for pipeline call topredict-translated-orfs
. See Issue #63 for more details.
- Unnecessary [analysis] package. Depending on how the package was installed, this sometimes resulted in the scripts not being found. See [Issue #56] (#56) for details.
- Handling of Stan model paths which include spaces. See Issue #36 for more details. This bug was reintroduced with the commit: "UPD resolve conflicts from dev for version 1.1"
- Backward-compatiblity syntax in
estimate-orf-bayes-factors
. See Issue #51 for more details.
__version__
and__version_info__
forrpbp
package
- Documentation to use pandoc for converting markdown to html
- Broken download link to example files
This is a rather significant update to the codebase and includes many small changes. Primarily, though, it addresses many installation and system issues.
In particular, thanks to Tonu Margus (@tmargus) for working through many of these issues.
- Installation instructions for anaconda
- Various installation issues related to calling
pip3
fromsetup.py
. Installation of private repositories is now handled via arequirements.txt
file.
- All uses of (py)bedtools
- Handling of Stan model paths which include spaces. In particular, this was a problem on OSX. See Issue #36 for more details.
The initial version, which implements everything in the paper. It was only used on debian-based systems.