LCAClassifier v2.0.5

Minor release to automatically download latest release of the UNITE database (v7 2016-01-31) and address installation problems due to deprecated version of buidlout's bootstrap.py script

LCAClassifier2.0.4

Bug fixes to avoid crash when aligned OTUs are missing from OTU table (can be casued using usearch as suggested in the UParse pipeline) as well as improved installation not automatically breaking with every new release of zc.buildout (which seems to be "a feature" of the later).

LCAClassifier2.0.3

Output is now written to a specific output directory, specified using option -o (by default "CREST_Results")

OTU-tables in delimeted text or BIOM format can now be imported together with alignments of representative sequences for each OTU (curated unique sequences or clusters thereof) to a CREST reference database. Composition for individual is determined from abundances in this table. Further, a taxonomically annotated OTU table or BIOM file is written to the output directory. When an OTU table is used, abundance data need not be included as annotation of sequence names. (If not, it still does however, using the default assumption that each Blast-file in XML format represents one dataset with the same name.)

Database alignments in gzipped (XML) format can now be used as input

Input from the blastn program part of NCBI's (relatively) new BLAST+ package can be used as an alternative to legacy Megablast. Recommeded settings are "-task Megablast -outfmt 5" (the later, ie XML format, being mandatory)