major upgrade - app uses vep api_version to look for available cache types (ensembl, refseq, merged) and disables any from selection if not available; results page displays api version and selected cache database; vep wrapper captures vep.pl stderr for display on results page; added vcf 'filter' col to results page; enabled HGVS notation function
moved forking config to config files - dev server faster without forks, deployment optimum with fork = 2; moved hires timer to immediately either side of vep script loop; deleted vep input file from /tmp if exists due to generation using append mode (in case previous run aborted before unlink) event; output format tweaks
added new counters for vep input rows and raw src_file rows; patched lab number regex to allow optional letter prefix; fixed test scripts to report vep input row count instead of 'accepted' rows
took vep_fields into CONSTANT for use in both 454 data & vcf data; put vcf data into %exon_data_ref for use in post-vep processing (as used in 454 data-files); adapted vep_location / ref_location comparison to handle insertions where end position < start position; moved counts & errors into separate 'data' hashref in results; completed adaptation of VEP for vcf data-files
added errors & alerts display to results output; checks all exon names in src data are in vep input data; updated test scripts with results output format
bugfix to _prediction_and_score() - needed to skip undef vars; colour-coded results if deleterious or damaging; truncated long (>30 char) variation results; re-ordered results cols for clarity