I have been curious about kit 14 and R10.4.1 flow cells. I will get to use some soon. Tonight, I watched Miten Jain from the University of California, Santa Cruz present at London Calling 2022 on “Human genome assembly and analysis using R10.4.1 , Kit 14, and duplex data.” Jain talked about “ultra long” sequencing (UL) as 100 kb+ read length. Jain explained that read accuracy continues to improve, with Guppy version updates going from 93 to 96% (R9.4.1 Guppy 4.2.2 to R9.4.1 Guppy 5.0.7). Jain noted that “UL-reads permit high-contiguity de novo assemblies.” UL reads help with high-accuracy assemblies. Jain announced R.9.4.1 ultra long read datasets for the Genome in a Bottle samples. Jain also ran kit 14 at high-speed (400 bps) and high-throughput. They were getting with two loads per flow cell >140 Gb per flow cell with less input requirement. With the low-speed (260 bps) runs with two loads, the flow cells still sequenced >120 Gb on the PromethION. R10.4.1 flow cells also produced better homopolymer sequencing. They also observed better SNP and variant calling. Jain concluded that ligation-based preps with kit 14 and R10.4.1 flow cells yielded 100 Gb per flow cell using sheared DNA, required less input without compromising throughput, and had high median alignments at both high and low-speed modes. Jain also explained that they are evaluating duplex reads too with this chemistry/flow cell. Jain concluded that ultra-long reads with high-throughput are possible on the PromethION (with PromethION flow cells).
