## 0.15.0 (2022-12-04)
- Fixed True/False highlighted column from 0.14.0 to include bindetect excel output as well.
- Added '--bias-pkl' option to ATACorrect (#174)

## 0.14.0 (2022-10-03)
- Added '--cmap' option in PlotHeatmap (#130)
- Added 'drop-chrom' in ATACorrect to gain more control over which chromosomes should be excluded (#142)
- Added a True/False "<cond1>_<cond2>_highlighted" column in bindetect_results.txt/.xlsx files. These columns mark the highlighted TFs for each volcano plot.
- Added check for empty '--TFBS' files in CreateNetwork

## 0.13.3 (2022-03-24)
- Fix of bug in utilities:match_lists leading to false list-to-list assignment in PlotChanges (#110)
- Fixed bug in creating gimmemotifs object
- Integrated checks related to gimmemotifs, pandas and python versions in ClusterMotifs
- Added check for fonts and 'gs' in PlotTracks
- Fixed bug for --track-off in TOBIAS ATACorrect

## 0.13.2 (2021-12-21)
- Fix for bug introduced in 0.13.1 - faulty reading of motifs from .meme files

## 0.13.1 (2021-12-06)
- Fix for #103 regarding BINDetect bound/unbound threshold
- Fix for #107 for MEME motifs with nsites=1 leading to bad_alloc

## 0.13.0 (2021-11-26)
- Added option '--norm-off' to BINDetect. Can be used in cases where the conditions are already normalized, or in case of issues in normalization.
- Fixed bug in BINDetect leading to errors when motif prefixes are the same, but in a different case (e.g. Jun/JUN). Fixed by making unique regardless of case.
- More checks in ATACorrect and BINDetect.

## 0.12.12 (2021-10-04)
- Fix for reading .meme files with float 'nsites' and consecutive motifs in file
- Better error message for missing 'gimmemotifs' import within ClusterMotifs
- Performance improvement of remove_duplicates and resolve_overlaps within regions utils
- Fix of bugs in ATACorrect (#94) related to setting bias threshold

## 0.12.11 (2021-06-21)
- Set cap on figure size in BINDetect for large number of motifs
- Fix for reading 1-bp .bed-files without "not proper bed format"-error in RegionList.from_bed
- Fix for reading .meme-files without trailing empty line

## 0.12.10 (2021-03-08)
- Within motifs utils: Added ylim parameter for visualization of motifs with low bitscores (from hschult) 
- Fixed bug in PlotHeatmap leading to error for out-of-bounds regions (#64)

## 0.12.9 (2021-02-16)
- Fixed bug in ATACorrect causing "uncorrected == corrected" for 1bp read .bam-files
- Fixed column matching in CreateNetwork so that target != source in the same file
- Minor change in regions utility to allow .bed-files ending in empty-string

## 0.12.8 (2021-02-02)
- Fix for similarity score error introduced in 0.12.7 (wrongly applied max instead of min)

## 0.12.7 (2021-02-01)
- Fixed error reading in meme.txt files (due to "----"-lines)
- Internal additions of motif-specific functions (e.g. GC-content, motifs from fasta)
- Capped motif similarities at 1 (influence rare cases where correlation < 0) 
- Added error for input '--peak'-files with varying number of columns in BINDetect

## 0.12.6 (2020-11-25)
- Fix for bug leading to keyerror when --peak-header did not contain "peak_chr"

## 0.12.5 (2020-11-25)
- Fix for error reading .meme-motifs with scientific notation (e.g. 1e-05)

## 0.12.4 (2020-11-02)
- Fix for color keyerror in BINDetect (volcano plotting)
- Fixed bug in BINDetect normalization leading to NA's in normalized values (and exit due to error). Fix now applies a robust sigmoid detection of norm factors per condition. NOTE: This introduces small changes in normalized values in comparison to versions 0.12+

## 0.12.3 (2020-10-23)
- Fix of windows /r line-end in "cluster_sites_by_overlap.py"

## 0.12.2 (2020-10-23)
- Fix to BINDetect bug occurring in case of duplicates in --motif input. Motifs are now given unique names if duplicated. 
- Change of "Transcription factor similarities" to "Transcription factor distance" in BINDetect.
- Added utility script "cluster_sites_by_overlap.py" to create a BINDetect-style dendrogram for overlapping .bed-file regions

## 0.12.1 (2020-09-28)
- Fixed bug in quantile normalization (from 0.12.0) leading to spurious normalization factors at high footprint values

## 0.12.0 (2020-09-09)
- Improvement of internal OneMotif/MotifList class structure for reading/writing/scanning
- Added '--output-peaks' to BINDetect to enable output of a subset of '--peak' regions
- Added '--output-txt' and calculation of agg-differences to PlotAggregate 
- Improved robustness of quantile normalization between conditions in BINDetect
- Fixed Nonetype error for peaks in ATACorrect
- Improved debug with added name of bigwig to bigwig reading error in BINDetect, ScoreBed and PlotAggregate
- Bugfixes to CreateNetwork for mismatches between --TFBS/--origin

## 0.11.7 (2020-09-07)
- Fix for BINDetect error when running with one input motif 
- Introduced dynamic fontsizes for PlotAggregate to deal with long TFBS/bigwig label names
- Bumped svist4get version to 1.2.24 due to missing font

## 0.11.6 (2020-07-23)
- Mess up of merge in 0.11.5. Same changes.

## 0.11.5 (2020-07-23)
- Fixed "out of bounds" bugs within BINDetect and PlotAggregate

## 0.11.4 (2020-07-20)
- Added noninteractive backend to all plot modules to fix tkinter error 

## 0.11.3 (2020-06-02)
- Fixed bug of undefined exit statement

## 0.11.2 (2020-05-11)
- Fixed bug when running BINDetect with very few peaks (no TFBS for many TF) and added warnings for this case
- Enabled running ATACorrect with bedToBam input - thanks to Surag Nair for 'read.infer_query_length()' fix!

## 0.11.1 (2020-05-04)
- Fixed error in parsing "patterns" for DownloadData
- Fixed errors for reading/writing bigwigs when pybigwig.numpy == 0
- Improved exceptions for bigwig/bam reading/writing

## 0.11.0 (2020-03-25)
- Removed 'gimmemotifs' from requirements and included check in MotifClust (only tool needing gimmemotifs) instead. This is a temporary hack solving various install/conda issues.
- Fixed smaller bug in bigwig writer for one-region output
- Added check for samtools in SubsampleBam
- Added @PG-line to FilterFragments output bam
- Added interactive html plots with motifs to BINDetect
- Adding --smooth function to PlotAggregate
- Fixed bug with overlapping figure labels in PlotAggregate
- Added .pickle output of ATACorrect bias object to use in other applications
- Fixed bugs in PlotTracks due to svist4get update; pinned =1.2.22 to setup.py
- Split internal structure of package into tools/utils/scripts
- Fixed bug in BINDetect cutting off last column of _bound/_unbound files.
- Added --naming to BINDetect
- Added DownloadData tool to enable download of test_data/ directory from s3
- Changed default --outdir's to be <toolname>_output
- Fixed one-off error in MaxPos
- Updated README with links to wiki
- Additional bugfixes

## 0.10.1 (2020-01-16)
- Fixed error reading parser and fonts for PlotTracks
- Restructure of motifclust-functions to be included in OneMotif and MotifList classes
- Smaller bugfixes

## 0.10.0 (2019-12-27)
- Added ClusterMotifs and FilterFragments to tools
- Restructured backend organization of parsers to speed up commandline 'TOBIAS'

## 0.9.0 (2019-11-21)
- Added PlotTracks to tools
- Fixed smaller bug with flank estimation in footprint calculation
- Fixed rescale to original sum in ATACorrect
- Added option to exclude outliers from aggregate plots

## 0.8.0 (2019-09-04)
- Added CreateNetwork to tools
- Smaller bugfixes

## 0.7.0 (2019-08-13)
- Updated the way p-values are calculated in BINDetect. Included 100-fold subsampling of background to estimate confidence interval.

## 0.6.4 (2019-08-07)
- Bug fix for error with regions very close to chromosome borders for ATACorrect and ScoreBigwig. 

## 0.6.3 (2019-07-16)
- Increased size of texts in BINDetect volcano plot and moved label into plot

## 0.6.2 (2019-06-19)
- Added --skip-excel option to skip excel overview per factor in BINDetect (can slow the run considerably for large output)
- Internal changes to BINDetect for better debug overview
- Fixed normalization for plotAggregate

## 0.6.1 (2019-06-05)
- Fixed motif cluster names to prefix C_
- Added --filter function to format_motifs to filter output motifs using names/ids

## 0.6.0 (2019-05-29)
- Added option to flip axes in PlotAggregate via "--signal-on-x"
- Changed all command-line arguments containing "_" to "-" (e.g. --regions_in to --regions-in) (but retaining both options internally)

## 0.5.3 (2019-05-28)
- Improved error messaging from file writers in bindetect

## 0.5.2 (2019-05-16)
- Bugfix for reading meme format files in MotifList().from_file()

## 0.5.1 (2019-05-15)
- Internal changes to OneMotif and MotifList classes
- Bindetect now takes directories/file(s) as input for --motifs

## 0.5.0 (2019-05-02)
- Added sum/mean/none scoring to ScoreBigwig as well as the option to get --absolute of input signal

## 0.4.1 (2019-04-29)
- Fixed weird "can't pickle SwigPyObject objects"-error in bindetect

## 0.4.0 (2019-04-29)
- Added --add_region_columns to TFBScan
- Renamed FootprintScores to ScoreBigwig
- Added normalization of input score distributions in BINDetect
- Added general warning for --cores > max number of processors when running multiprocessing
- Updated PlotChanges to also plot mean_score and percent_bound as well as overview per motif cluster
- Added MOODS-python as dependency through pip

## 0.3.1 (2019-04-10)
- Split motif ids between name/motif_id/output_prefix
