Biocore 0.3.1 is released

by Ketil Malde; January 7, 2013

After discussion on the biohaskell mailing list, version 0.3.1 of biocore is out there. New features are more class instances, the various data types are now Monoids (meaning you can easily concatenate them), and the somewhat vaguely defined seqlabel is replaced by seqid (returning a single word identifying a sequence) and seqheader (returning a FASTA-style header containing id and comments).

Although it was discussed at lenght, there is no slicing and dicing functionality included. My personal feeling is that this is better achieved by users implementing ListLike or some similar set of functions. I’m also hesitant to export by default names that clash with the Prelude (and hosts of other modules, the authors of which have no such reservations), and finally, slicing isn’t really well defined for some sequence types - for instance the SFF type used by 454 sequencers are really a vector of “flow values”, and you can’t really cut the resulting nucleotide sequence up arbitrarily.

The astute reader will have noticed there is also a version 0.3 available, the main difference is that the seqlabel function is now back in the BioSeq class (but still deprecated, of course), and default implementations have been added, easing the transition for libraries who instantiate the class by defining seqlabel. Also, the definitions of toFasta and toFastq were updated to use the new functions, so that toFasta now outputs the full header.

Oh, and biosff and the included flower program have been updated (version 0.3.2) to work with the new version of biocore. The rest of the stuff I tested seemed to work without modification (but of course, do let me know if I’m mistaken!)

comments powered by Disqus
Feedback? Please email