GitLab

now gets protein_id vai previous dump's embl file; includes names, descriptions, Dros orhtologues; deals with bacterial genes; has comments

use a straight_join when joining to repeat_consensus on repeat_type, because the mysql query optimiser gets it wrong

removed fetch_all_by_Slice_and_type method, extended fetch_all_by_Slice method to take additional type arg

cleaned up debugging output, made silencable; StatCodes can be logged to file for later analysis/retrieval

fix: dbh->{Kids} is only decremented after DESTROY, so disconnect needs to be called by destroy if Kids == 1

DBConnection::disconnect_if_idle now uses $dbh->{Kids} instead of $dbh->{ActiveKids} to prevent problems with multiple prepared statement handles

Transcript::add_Exon now throws if exon could not be added b/c of overlap with other exon, rather than silently doing nothing

*do not create translation if less than 3 nucl. left in CDS

bug fix: still adjust coords for exon that is wholly consumed by introduced frameshift intron b/c coords are used later for transcript splitting.