Thread by @AlbertVilella

In #NextGenerationSequencing news, we look at the clinical uptake of different NGS technologies after the recent updates announced in the last few months:

NGS has been dominated by Illumina with their SBS short-reads technology, with reads of up to 2x150 base pairs available in most configurations, although 2x300 are also possible at a higher cost per Gb.

If we could roll back time in the last 15-20 years of NGS history, we would ideally have had a long-reads technology, or better yet, a length agnostic technology, be the dominant player, as most applications, both in Life Sciences Research and Clinical, would have benefited

from reads longer than the ones Illumina produces. As such technologies, long-read or length agnostic evolve, we should see a segmentation of the Clinical market where some applications are done exclusively via long-reads, whereas others remain as Illumina SBS applications.

One example of the later could be Liquid Biopsy, e.g. for cancer screening, by sequencing of the cell-free DNA that circulates in our blood stream. It just so happens that the length of the majority of such DNA fragments is about ~166bp, which is well suited for short-reads.

The market for Liquid Biopsy, and particularly for cancer screening, could grow up to be a $10B, or $30B or even $50B market, depending on when we place the landmark and how we define it. But taking together cancer screening and MRD, Non-Invasive Prenatal Testing (NIPT) and

similar segments, there will be still a big appetite for short-reads. The Grail takeover merger is an examle of the strategic importance that Illumina places on Liquid Biopsy diagnostics. Will they be allow to vertically integrate in such a way, and thus punish the Guardant

Health and Exact Sciences of the cancer screening world: probably not.
Then what else is amenable to long-read technologies in the Clinical space? Literally everything else.

An example here is the genomic sequencing of new born babies. At the moment this is restricted to neonates that present a phenotype that may be triggered by a genetic condition, but if we project this into the future, there could be a near future where every single new-born has

their genome sequenced at birth, and this becomes their first element of their electronic health records. We've seen examples recently of how good genome sequencing and assembly can become with long-read technologies, such as PacBio HiFi reads, and even more complete with the use

of Oxford Nanopore Ultralong read sequencing, where each read can reach length of over a megabase, which is a couple of orders of magnitude longer than the standard genomic DNA prep, e.g. the transposase-based prep of genomic DNA that Oxford Nanopore offers.

A combination of those two would deliver a complete assembly, even crossing over the centromeres, of a human genome, at which point, this initial EHR becomes a reference on top of which other layers of multi-omic information can be layered over.

The use of these technologies is becoming more mainstream, and recent advances both by PacBio with their new Revio system, and Oxford Nanopore with their Kit14 and more recently with the preview of their "stereo duplex" and 1000-base-per-second upgrades will mean

that the advantage that Illumina used to have, in sheer throughput and availability of their systems, is narrowing. Illumina's revenues are also not growing as they used to, whereas Oxford Nanopore has had their last two years of excellent growth in revenue, with the last one

sustaining good revenue even after discounting the COVID-19 variant surveillance revenue that peaked two years ago. The fact that Oxford Nanopore is very attractive in terms of CAPEX and can be widely deployed in a range of setups is an advantage for the growth into clinical use.

As a comparison, the newly announced Revio instrument by PacBio will deliver a near 30x HiFi genome for $1000, with the instrument at a price of $779,000 and capable of producing 1,300 genomes per year.

Oxford Nanopore has been selling instrument that already outmatch the Revio in throughput, with the P48 capable of producing ten times the number of genomes per year. The new P2 instrument is also a very attractive proposition, with a starting price of $10,500, it is ideal for

setups where the NGS doesn't need to be a centralised core factory setting, and the instrument can remain available as the samples come in for sequencing, e.g. in a small hospital or health centre.

The new Revio at $1000 per HiFi genome, and the Oxford Nanopore Kit14 Promethion flowcells are already close enough to the Illumina short-read equivalent that many clinical applications can switch to long-reads without a significant impact in their operations.

In fact, the new P2 instrument by Oxford Nanopore opens new doors for Clinical applications in settings where neither a large Illumina or PacBio instrument could be placed so far. This is a new boost for growth in the NGS space where other instrument types are already saturated.

There are aspects of familiarity, especially in sample extraction and library preparation, where these long-read technologies will have to continue educating the customer base. The fact that Illumina SBS is limited to short-reads meant it didn't matter if the input source of DNA

got damaged, i.e. fragmented, during extraction and preparation, which is not the case with long-read technologies. Here, the better we purify the DNA, the better the read-out we will get. Considering that both PacBio and Oxford Nanopore can read 5mC out of native DNA, means we

also get the epigenetic information on top of the genetic read-out of a sample.
These are exciting times for NGS, and if you want to read more about them, follow this account and go to bit.ly/ngs-slides for a summary of the who is who of NGS.