Nanopore adaptive sequencing for mixed samples, whole exome capture and targeted panels
Matt introduced the Read Until adaptive sampling toolkit. With this, >90% of reads were called, mapped, and decided within 1.2 sec of read start.
Adaptive sampling enriches for organisms in metagenomic samples: they achieved a target of 40x depth for all 8 genomes in the Zymo mock community, incl. S. cerevisiae at 2% prevalence.
Matt: I proposed ‘A Crazy Experiment’: exon sequencing of the human genome with adaptive sampling. This involved 25,600 targets in ~10,000 genes, on all odd-numbered c’somes. It worked!
Next, targeting ‘useful’ areas of the genome: they achieved 31x enrichment of 717 cancer-related targets, in a ~10 Gb run (~3x would be expected without enrichment).
Matt concluded that selective sequencing ‘really does work’, and you don’t need a large GPU to do it!
