🔗 https://doi.org/10.1093/bioinformatics/btaf416
💻️ rdeval code: https://github.com/vgl-hub/rdeval (3/3)

Huge thanks to the whole team, especially @hillermich.bsky.social @cmazzoni.bsky.social @astridboehne.bsky.social Sylke Winkler, Carola Greve,
@yanischrysost.bsky.social Larissa Arantes, Tom Brown, Charlotte Gerheim, Tilman Schell !!!

Based on our and another study, @pacbio.bsky.social has now oficially adopted polymerase C (KOD Xtreme™ Hot Start DNA polymerase) in their new Ampli-Fi input protocol:
www.pacb.com/blog/new-amp...

Applying polymerase A/B and C read amplification to two mollusc and one arthropod sample, we saw that assembly contiguity was either higher or comparable with polymerase C reads, and assemblies were always more contigous when combining all libraries, probably due to complementary read drop-outs.

Combining reads of polymerase A/B and C with HiC reads, we were able to assemble the maned sloth genome to chromosome level with a contig N50 of ~4.75 Mb, a 10x increase compared to polymerase A/B reads alone.

The assemblies 3.1 Gb size also surpasses the previous 500 Mb protocol size limit.

We then used a different polymerase "C" to repeat amplification and sequencing, increasing contig N50 and compleasm/TOGA gene completness dramatically.

Aligning reads of B. torquatus to another sloth species, we saw obvious read drop-outs that polymerase C reads could cover:

We therefore switched to the ultra-low input protocol, which contains a PCR amplification step of input DNA, employing two different polymerases we call A/B.

This way, we could sequence the maned sloth Bradypus torquatus to a coverage of 45X.

However, contig N50 of the assembly was below 1 Mb: