[~jprocter] says: Finally, for concrete use, it could be really cool to have a simple BAM read browser based on the overview window. Rather than show actual reads, the view would show a pileup based on windowed aligned read counts on a chromosome. That, combined with a table of read counts per annotated gene, would make it easy for someone to drill down. Of course, for now, we could also do exactly the same as for 'load VCF' - use any reference genome coordinates that can be discovered on the alignment to query the BAM file for relevant reads. Any discovered could then be added as subgroups for each sequence in the alignment.