Release 2.3.5

Release version for Conda integration

Release 2.3.3

Changes:

• Added option for biobloomcategorizer to use dust filtering to prevent low complexity sequences from classifying
• Added option for biobloomcategorizer to have the ability to switch between classification scoring methods
• binomial scoring maintains better classification sensitivity on average given reads with variable length and sequence quality
• Updated ntHash to improve hashing speed (new version is up to 4 times faster in hashing)
• Other refactoring such as decoupling miBF code from repository to btl_bloomfilter repository
• Various bug fixes

Release 2.3.2

Improved miBF construction. Changes will minimize ID bias and faster constuction.

Will use roughly twice the memory as before only during miBF construction.

Release 2.3.1

Fixes to ntHash2 for spaced seed hashing. Now possible to use miBF with spaced seeds correctly.

Fixes to paired mode with miBF when input has non-ATCG characters.

Release 2.3.0

Version checks for older bloom filters added to prevent accidental use of older filters.

Multi-index Bloom Filters are bugged in this version due to a bug in nthash2.

Release 2.2.0

Multi-index Bloom Filter support. See Doc/MultiIndex.md for details.

Fixes to biobloomcategorizer

ntHash is updated in this version so older Bloom filters are not compatible.
Multi-index Bloom Filters are bugged in this version due to a bug in nthash2.

Release 2.1.2b

Release for Official Multi-index Bloom Filter support.

Fixes:
Header output now includes comment.

New executables:
biobloommimaker
biobloommicategorizer

See Doc/MultiIndex.md for details.

There are known bugs introduced into biobloomcategorizer in this release

Release 2.1.1

add missing includes

Release 2.1.0

A Public Release of BBT using ntHash and Experimental multi-index Bloom filter (miBF).

The addition of ntHash and other optimizations in this version should half the runtime.

See help dialog as the interface has slightly changed with some added features.

BBT no longer supports files other than fasta or fastq (in either gz or non-gz form) to streamline I/O.

Due to the use of a new hash function old filters are not compatible with this new version.

New release /w Hotfix

General:

• General Bugfixes
• -s and -r now accepts an integer that represents the number of contiguous matching bases rather than just based on score

Biobloommaker:

• no longer requires and fai file to create filters
• can now be threaded
• Optimized I/O for faster filter creation performance
• -s can now be used with progressive mode