Replace original filtering with ensembl canonical tag

[kkuret]

Is your feature request related to a problem? Please describe

Describe the solution you'd like

Switch from basic tag and TSL- based filtering of the GTF before calling ic-count mini, to a simpler filtering that only keeps for each gene the transcript labelled as Ensembl_canonical

Describe alternatives you've considered

Additional context

[iraiosub]

Ensembl_canonical is missing from many annotations. Updated plan is: We are instead simplifying by filtering the GTF based on either user input transcripts (one representative transcript per gene) if provided; if not provided the pipeline by default detects the longest transcript from the original input GTF (with the following hierarchy: longest CDS > longest tx (sum of all exons)) and keeps one representative (longest) transcript per gene. see issue #141