Bugfix 1.1.3

1. Fixed methratio.py to handle "no-action" CT SNP
2. Added test to verify this works in the future

Fixes

#28

Bugfix 1.1.2

• Refcache now falls back to inefficient disk reads for out of order fetches
• Added action to build conda packages on releases
• For experiments with extremely deep sequencing, the -I flag was added to use unsigned integers instead of unsigned shorts counts. Please note that this will use twice as much memory.
• Updated sam2bam.sh for recent samtools functionality
• Updated requirements and installation instructions

Fixes

• #22
• #20
• #10
• #17

Bugfix 1.1.1

1. Fixed methylation ratio calculation in methdiff.py:
   • Methylation ratio now uses min(C, eff_CT)/eff_CT to be consistant with methratio.py
   • span value now uses floating point numbers to stop accidental truncation
2. Fixed LICENSE attribution
3. Conda target will now correctly build with conda libraries and compilers

Release 1.1.0 - Fixes and features

Changes

1. The original version of methratio.py did not trim overlapping reads from BSP input. This issue has been fixed.
2. Implemented memory limits to methratio.py (-M), which defaults to 90% of available memory
3. Implemented processor limits to methratio.py (-P), which defaults to ALL cores
4. Added the option to return full methylation context (CAG instead of CHG) with methratio.py
5. Test suite now compares all possible output formats for single and paired reads

Fixes

1. Fixes #8

Notes

The original version of methratio.py counted overlapping regions of reads twice when using BSP input. We recommend re-running all methylation extraction if you utilize the BSP format.

Initial Release of BSMAPz 1.0

This is the first stable release of BSMAPz

Changes from BSMAPv2.90

1. Patched samtools and main.cpp to for compatibility on OSX
2. Updated README
3. Patched methratio.py for compatibility with modern samtools
4. Added test target in Makefile
5. Fixed parallel methratio.py bug
6. Validated methratio.py for all input types
7. methratio.py will automatically process chromosomes in parallel and use less memory