Tonight I watched the entire Targeted sequencing showcase from London Calling 2023. The first panelist was Lukas Wellguny, a Ph.D. student from EMBL-EBI in the UK. Their presentation title was “Dynamic, adaptive sampling during nanopore sequencing using Bayesian experimental design.” They take data from the sequencer and incorporate a model to derive uncertainty scores at each site of the genome. Then, the program uses adaptive sampling to fill in the gaps or target variants. Wellguny tested the algorithm with a mock community to demonstrate more even sequencing. A reference-free dynamic adaptive sampling system enriches for rare species in their experiments. The metagenome assembled genomes (MAGs) from these enrichment experiments were more complete than without the adaptive sampling and algorithm. Next, Romain Boidot from the Centre Georges-François Leclerc in France presented a short session entitled “The potential use of nanopore sequencing for routine molecular diagnosis of germline cancer predisposition.” They began with a description of the need for detecting unique regions of the genome for molecular diagnostics. The objective of their study was to determine the ability to detect target genes with a streamlined bed file and efficient use of the flow cell. Boidot shared data on the impact of enrichment system. Next, Thidathip Wongsurwat from the Siriraj Long-read Lab at Mahidol University in Thailand spoke about “Rapid identification of bacterial invasion in the amniotic cavity: full-length 16S or adaptive sampling?” Their team compared cultivation, Sanger sequencing, and Nanopore sequencing approaches for species identification of infection of amniotic fluid. They used a GridION and pus from patients. It takes technicians 80 min to prepare the library using the LSK 114 library prep system. Hours of sequencing helped identify hard-to-culture bacterial organisms. During the question and answer session, panelists were asked about what they would like Oxford Nanopore Technologies (ONT) to innovate on and improve. Panelists spoke about improvements in adaptive sampling and multiplexing samples were desired. While one flow cell cell is needed for adaptive sampling, Wongsurwat noted that their team was able to obtain species information within hours. One audience member asked about the use of adaptive sampling for direct RNA sequencing. Audience members requested information about adaptive sampling with PromethION flow cells. The last question was challenged panelists to think about additional applications and opportunities. This session made me think about what information we have and how ONT can be used. Rewatching the complete session emphasized how adaptive sampling will improve workflows and, in turn, imrpove efficiency of programs and flow cell use.
