RELEASE.md
#### Version 1.0.0
Released October 20th 2014
#####Features:
1. Identification of SV breakpoints.
2. Identification of soft-clipped consensus.
3. Provides the end of breakpoints using supplement alignments.
4. Provides genotypes and genotype likelihoods for multiple individuals.
5. Multi-threading enabled, supported by OpenMP.
#####Known Issues:
1. For large population calling WHAM can hit the Ulimit.
2. For deep regions WHAM has considerable slowdown.
3. Memory becomes an issue for deep regions.
#####Future efforts:
1. Improvements to speed and memory usage.
2. Improve WHAM's break-point joining.
3. Improve the LRT for association testing.
#### Version 1.1.0
Released October 24th 2014
#####Features:
1. WHAM now only reports SVs where the consensus is longer than 10bp.
2. WHAM can now be built in debug mode, which prints all critical data structures
```
make debug
```
#####Known Issues:
Same as Version 1.0.0
#####Future efforts:
Same as Version 1.0.0
#####Bug fixes:
1. Now using discordant mate-pairs correctly.
2. Major improvements to accuracy on simulated data
3. More SVs have paired breakpoints
#### Version 1.2.0
Released October 31th 2014 - happy Halloween!
#####Features:
1. Increased overall accuracy of SV calls
2. Added attributes to VCF file for SV type classification
3. Added utils/classifier_parse.py for SV classification
4. Added functionality to readPileUp Class
#####Known Issues:
Same as Version 1.1.0
#####Future efforts:
Same as Version 1.1.0
#####Bug fixes:
#### Version 1.3.0
Released Nov 4th - happy voting!
#####Features:
1. Increased overall accuracy of SV calls.
Large calls and translocations require more support.
2. Added another attribute for the classifier.
#####Future efforts:
1. Validation on published data sets.
#####Bug fixes:
#### Version 1.4.0
Released Nov 12th - happy Veterans Day -1
### Features:
1. Added deltaAfWham for pooled sequencing experiments.
2. Classifier now prints cross-validation information.
#####Future efforts:
1. Adding subsampling to the genotype model.
#####Bug fixes:
1. High depths were causing the genotype likelihood model
to underflow. This is fixed. Only the first 1k reads
are considered for a genotype call.
2. Fixes to the VCF header again
#### Version 1.5.0
Released Nov 17th
### Features:
### Bug Fixes:
1. A pileup bug was resulting in bad genotype calls.
#####Future efforts:
Futher efforts to improve genotyping.
#### Version 1.6.0
Released Nov 30th
### Features:
1. Added mate-pair end-point clustering.
2. Several INFO and GENOTYPE fields were changed.
### Bug Fixes:
None
#####Future efforts:
More benchmarking.
#### Version 1.7.0
### Features:
1. Added average mapping quality (MQ)
2. Fraction of reads with quality lower than 50
3. Filtering out reads with mapping quality below 40 (better specificity)
4. Set several cutoffs based on CHM1 data
### Bug Fixes:
None.