- Updated the logic of using
only_oncogenes
for filtering.
- Fixed the data loading for oncogene of mouse.
- Updated Sequenza workflow.
- Fixed the re-handling of errored seqz and facets runs.
- Supported analysis workflow from FACETS and Sequenza.
- Mouse genome is enabled based on the two implemented workflows above.
- The first public and stable release.
- Limited
xgboost
version lower than 1.6
as it will not keep some key info in .rds
file.
- Enhance the
getGeneSummary()
and getCytobandSummary()
methods to return mutation matrix.
- This version is not compatible with previous versions, as the analysis and visualization functions are moved to an independent package ‘gcaputils’.
- Added some utils functions and visualization functions.
- Updated initial setting and CLI.
This version has been discarded from git history.
- Supported a
NA
passing as tightness to remove the use of TCGA blood summary data as a more strict threshold for circular amplicon.
- Added
fCNA$subset()
method.
- Added
gcap.plotDistribution()
function.
- Added
gcap.plotForest()
.
- Added
gcap.plotKMcurve()
.
- Added
gcap.plotProfile()
.
- Added method
convertGeneID()
to fCNA
class.
- Set a default value for
pdata
option in fCNA$new()
.
- Supported
gcap
as main command, and previous two commands as subcommands if GetoptLong
version >=1.1.0
. Note: not test yet.
- Added options
tightness
and gap_cn
.
- Renamed
gcap-wes.R
script to gcap-bam.R
.
- Handled void result.
- Updated the background copy number reference and criterion judging a amplicon (#22).
- Designed and implemented a class
fCNA
used for storing the workflow key outputs and downstream analysis and visualization.
- Provided a function
convertID()
to convert gene IDs.
- Optimized the output summary.
- Re-constructed scoring and workflow output.
- Added
use_best_ntreelimit
in gcap.runPrediction()
to control the ntree setting. When it is FALSE
, we use a custom processing to obtain a more conservative tree number.
- Added
deploy()
to auto-deploy the CLI to /usr/loca/bin
.
- Added easy-to-use CLI in
inst
directory.
- Filled
NA
s to input when age and gender are not available.
- Automatically appended logs to specific directory with
rappdirs::app_dir("gcap", "ShixiangWang")
. Users can obtain log path and cat log info with gcap:::get_log_file()
and gcap:::cat_log_file()
for debugging. (#14)
- Supported XGB54 model in workflows. (#13)
- After exploration, we found our stepwise model outperform MBO tuned model. So the models for predicting circular target have been limited to 3.
-
gcap.ASCNworkflow()
now supports input with only total integer copy number, like the result from ABSOLUTE software (also DoAbsolute).
- Added stepwised model for circle target.
- Changed the way how to select model and run prediction.
-
custom_model
in gcap.runPrediction()
has been changed to model
. This is inconsistent with version below v0.4.
- Updated scoring for supporting different thresholds.
- Wrapped ASCAT workflow in
tryCatch()
to avoid abnormal failure.
- Fixed input
extra_info
sample (order) issues.
- Fixed the issue about rendering wrong data files when skipping existing ASCAT calling.
- Updated models for prediction.
- Removed
prob_cutoff
setting in workflows. Directly use prob 0.1, 0.5 and 0.9 for cutting low, medium and high quality amplicon.
- Implemented an alternative workflow from allele specific copy number data to final result files. (#5)
- Implemented basic workflow from BAM files to result files.
- Added a
NEWS.md
file to track changes to the package.