Version 1.5.0

This version uses Python 3.7 and up, and it should work will the newest versions of all the dependencies.

Version 1.4.2

Newer versions of lifelines raise a ValueError when the algorithm can't converge. This version of genipe now catches this error and warns the user (instead of quitting).

Version 1.4.1

Fixed an issue when launching jobs with DRMAA and the "start time" or "end time" of the job wasn't reported. If "start time" is missing, it is replaced by the "launch time". If the "end time" is missing, it is replaced by the current time when DRMAA returned the job as completed.

This version of genipe should now work with the most recent versions of the different python modules. It has been tested with the following versions:

Module	Minimal tested version	Tested up to
numpy	1.11.3	1.14.3
pandas	0.19.2	0.23.0
lifelines	0.12.0	0.14.3
statsmodels	0.8.0	0.9.0

Version 1.4.0

This version of genipe should work with the most recent versions of the different python modules. It has been tested with the following versions:

Module	Minimal tested version	Tested up to
numpy	1.11.3	1.13.3
pandas	0.19.2	0.21.0
lifelines	0.12.0	0.12.0

Version 1.3.3

The tutorial script was updated:

• Checks binaries in the system path before downloading them.
• Updated link to download Plink.

Because of the different module's deprecation, this release should work only with the following module's version. Newer versions wouldn't work with this release of genipe.

• pandas version 0.19
• lifelines version 0.9

Version 1.3.2

This release includes the following changes:

• Compatibility with statsmodels version 0.8.0.
• Conda package for Python 3.6.

Version 1.3.1

This release includes the following changes:

• When launching the pipeline, using --chrom autosomes will now launch the analysis on all autosomes (from chromosomes 1 to 22).
• The tutorial script (genipe-tutorial) now performs the analysis on the autosomes.

Version 1.3.0

Changes to the main imputation pipeline include:

• Can now impute only a subset of chromosomes (the --chrom option).
• Possibility to further customize SHAPEIT (the --shapeit-extra option).
• Possibility to further customize IMPUTE2 (the --impute2-extra option).
• Chromosome X (both pseudo- and non-pseudo-autosomal regions) are now imputable.

Changes to the statistical tool include:

• Mixed linear model is now optimized to significantly reduce computation time.
• Better handling of categorical variables in survival analysis (CoxPH) (the --categorical option).
• Linear analysis now returns the adjusted r-squared value.

Changes to the extraction tool include:

• Addition of the LGEN output format (the --long option, available for the calls and dosage format).
• Addition of the BED output format (the --format bed option).

The compatibility with Python 3.3 was removed (since pandas versions are too old in anaconda for this version).

Version 1.2.3

Changes include:

• A standalone script that downloads reference files, genotype files and binaries to help the user execute the main pipeline tutorial (be sure to check the binaries licences (see here for more information).
• Better documentation.

Version 1.2.2

Changes include:

• Automatic report
  • creates a bar plot instead of a pie chart for better frequency comparison
  • minor amelioration in the template (for file names)
  • option (--report-background) to change the content of the background section
• impute2-extractor
  • possible to only index impute2 files using --index
  • when extracting in impute2 format, the companion files are also processed
• corrected some tests that failed depending of system used