James Bonfield

Good point - I forgot I can just transfer things. :) Edit: now transferred.

I should also raise CRAM here. BAM we know only stores data in nibbles and it's simply impossible for it to represent U as U. It *must* be T and...

> Since @Psy-Fer mentioned that [hybrid cases](https://genomic.social/@Psy_Fer_/113396324140980234) (reads containing both U and T) would usefully be representable, I'm perhaps leaning more towards a tagged field rather than a flag bit....

I'm not sure practically speaking it matters much about library construction here for SAM/BAM et al. The adapters would get trimmed away anyway pre alignment. If we consider an unaligned...

That's a good point. However fundamentally it breaks BAM. BAM simply cannot encode U. So much goes via BAM (including htslib's core in-memory format due to where it came from),...

I'm not sure that adding a badge here is the correct thing do, but we do plan on expanding the web site to report a more wider list of options...

I would assume we simply have to interpret GL as the likelihood that those alleles are the correct ones, and if we want to be sure that the phasing is...

You may wish to try `bcftools mpileup -X pacbio-ccs` to select options optimised for HiFi data. There's variability of course between organisms and sequencing runs, but I would expect it...

That's a good point. CRAM v1 didn't have RR. It was added for v2 so we could encode data without needing a reference; useful for unmapped data and also sometimes...

I had a hunt through the commit history and it was there on day 1 (well day 1 of commit history) in 2013 for VCF 4.1. I was optimistically hoping...