Changes/fixes have been implemented and are available at the production version of the TNRS. The following items have been fixed or added:
Tab conversion to whitespace
Escaping of trailing and leading whitespace
Escaping special characters
Removal of backslash on accented characters
Documentation for flags is updated
Flags have hover over with ability to click for more details
Score calculation changes (for clarity)
Family fuzzy matching
Database schema is being updated and populated with NCBI taxonomic data. This should be loaded within the next week, however will require testing and modification to deal with homonym resolution.
UI work is underway to support the following:
Selection of one or many sources, including ranking of sources for matching.
Selection of one source for classification (NCBI varies from TROPICOS as NCBI is more clade based)
Allowing for user selection of the match threshold desired.
Filtering for a particular category of plant species (mosses, angiosperms, etc)
Algorithm updates are being made to adjust for some misleading matches within a genus. We identified an issue where a 100% correct genus name was submitted with an invalid species. The algorithm matched the entire string to the wrong genus as the highest match (because it fell within the edit distance). This is being remedied to apply more weight to a correct genus match.
Expected delivery date for completed product is early/mid-November.
PHLAWD
Complete set of PHLAWD tools installed on Lonestar
Installed mafft, muscle, quicktree, phyutility, sqlite for both Intel and gcc compilers
Complete set of PHLAWD tools ready to be installed on Ranger
Same set as above, we just have to push them out
RAxML
Continuing to run 75k and 100k cases for Alexis and Stephen Smith
Ran a small case with Fernando Izquierdo's perpetually updated tree RAxML work flow