minor fixes - blast dbv5 compatible

Updates

• build-in blast_regexp now captures even non-standard accessions present in refseq and updated to capture new pdb chains with up to 4 character chains id's
• add support and recommend blast 2.9 due to update of blastdb format (dbv5 - lower then 2.8.1 not compatible with now used ncbi databases)

Html output updates

• colour scheme change
• message for uncertain extended hits
• added original HSP position indication to the sequence viewer (with "--show_HSP" flag)

New "--entrez" option for downloading needed sequences directly from rboAnalyzer pipeline

• suitable for one-time analysis
• works only if the ids captured by the "blast_regexp" are accession.version id's present in the NCBI nucleotide database
• valid email address is required with this option
• if provided the "-db" option is not used

Html output update and patch for very long hits

• change help appearance in html output
• remove unused filed 'bstrand' from csv output
• patch for very long hits (anchors) (locarna can't process more then 520 anchor points)

Patch for pdb entry in BLAST xml output

Fixes case of unrecognised sequence accession from pdb resulting from mismatch between accession in BLAST database and reported accession in BLAST xml output.

Patch for new NCBI SeqViewer

Fixes the code for new version of NCBI Sequence viewer which didn't work correctly after update to 3.31.

New stable release

• All system calls wrapper in try except preventing potential errors from propagating.
• Incorrectly installed dependency (i.e. appears to be there but fails with real data) will not crash computation
• --filter_by_eval and --filter_by_bits now accept intervals (i.e. >=5, <200)
• Each rboAnalyzer computation is now saved to enable faster computation in case of repeated calls with e.g. different filtering options, different prediction methods. The data are saved in the same path as the BLAST output which is analysed, the file name is constructed as follows: BO_file_name + .r- + hash string. The hash string corresponds to the arguments which would change the extended sequences.
• force to have some output requested on command-line
• removed --keep_all argument
• rboAnalyzer does not ask about continuing the interrupted computation. It assumes that you want to continue. The parameters which would influence the result are checked.

bugfix release

• fix error when uncertain covariance model
• catch all exceptions, even those not primarily caused by rboAnalyzer

stable

• Renaming prediction methods.

• Adds ability to continue with interrupted computations.

• Logging redesign.

• Require Biopython < 1.73 (NCBIXML parser differs in 1.74)

• Documentation updates.

• Example directory attached for easy download.

beta

Changes

• Updated documentation
• Added options
• reworked code

update docs

test conda installation

Updated output

Fixes:

• repeated location in version description with method "joined"
• inconsistent reports with methods "simple" and "joined"
• location one of in certain cases
• javascript functionality in html output

Enhancments:

• javascript function now more robust