Geo Pertea

It's still puzzling that the `--fr` option did not seem to work properly there. I'd be curious if `--rf` would've done something else there. I guess it is possible that...

Is this a trick question? :D Well, the StringTie protocol (if you check the papers or the web page) was based specifically on HISAT2 alignments. So our very understandably biased...

Ah, our annoying FTP server is acting up again.. I switched those downloads to http, please try reloading the hisat2 web page and click the donwload links again.. (only software,...

Keep in mind a few notes when using HISAT2 (and use the example StringTie2 protocol for guidance, alongside the HISAT2 manual). - if you want to use reference annotation as...

No, please don't merge the 4 BAM files (or their stringtie outputs), it's actually messing up the results instead of helping.. You should pick only one aligner/strategy -- you could...

Yes, tough call.. that STAR.fr run does indeed have the best sensitivity at intron/chain level, though precision (Sp) is not that great.. Quite strange that some of the Sp values...

I also noticed in some recent simulated experiments that, with the default parameters, aligning to a genome index without any junction data, STAR has a much lower sensitivity/recovery rate for...

Admittedly gffread has some shortcomings when it comes to preserving the original attributes, and these shortcomings are especially obvious for non-transcripts (-O) *and* for GTF output (-T). The -T option...

Hm, there is enough trouble with the fact that GTF/GFF3 formats are way *too loose* and now some annotators are even adding stuff like this which lacks the commonly expected...

Ouch.. Gene features in GFF3 make sense but in GTF they are an abomination :), I did not know that Ensembl started committing this atrocity too (so far I thought...