TTP's case studies page has a large number of articles on a wide variety of topics; as a whole the firm covers a wide range of technology fields far beyond genomics or medicine. TTP is also the origin of sptLabtech, which was spun out in 2018 and makes a number of instruments such as the Mosquito and Firefly liquid handlers.
The tagline on Puckdx pretty much describes it: robot in a coffee pod. Reagents are packed into a small pod-like device that is then served by a central pipettor. The hardware portion could easily fit in a large suitcase, and TTP assures me that it actually has quite a bit of empty space and could be slimmed down further. PCR in the device can be fast: 30 cycles in 7 minutes.
Diasorin licensed Puckdx for human in vitro diagnostics, but TTP held out other uses (such as animal or plant agriculture) as well sample preparation. What TTP displayed at ASHG is a reagent load and protocol which takes crude cell lysates as inputs and produces in just under 2 hours an Oxford Nanopore ligation library. So purification and all the steps in a ligation library, such as bead cleanups, in a single device.
The end user is encumbered with little hands-on work, though the tricky final flowcell priming and loading is left to them. I'm sure TTP would enjoy engineering loading into the device as well, though early ONT decisions on the flowcell design don't make that easy; the latch on a MinION flowcell is annoying enough to work with human hands, let alone robotic actuators.
That does beget a question: is a single ligation library have an economically viable application for field-deployable prep instruments? Anyone in this space is pioneering markets that don't exist, and may not exist.
This was one of the challenges ONT's own field-deployable VolTRAX instrument as well as the bulky ElysION is that both ran only rapid preps, which are already very few pipetting steps. VolTRAX also still required manual loading. So why buy a bunch of proprietary kit to automate a simple 10 minute protocol?
The TTP device does move farther upstream, automating DNA extractions as well as library preparation, which may be a significant value-add. As discussed in my recent piece on seqWell's ONT prep, ligation libraries have higher yields than the standard ONT rapid prep, though seqWell says they close the gap with their added purification step. One application TTP mentioned is the very cool "yes, it is brain surgery" work being pioneered in the UK - giving feedback to a surgeon while they have a patient on the operating table as to the precise tumor the patient has and whether that warrants conservative or aggressive removal of tissue margins.
Perhaps the better demonstrated nanopore sequencing markets with interest in portable devices are PCR amplicons and particularly PCR coupled with reverse transcription to detect RNA genome respiratory viruses. The latter has the downside that in general the viral tiling strategies require splitting the sample at some point, typically after reverse transcription, so that complete genome coverage can be achieved with overlapping amplicons. Swift Biosciences had a kit that covered most of the genome with a single tube amplification by using very carefully designed primers, and there are also approaches that use one universal priming site introduced during second strand synthesis combined with specific primers to enrich for the target genome. But split-sample tiling schemes such as ARTIC dominated the pandemic sequencing and are still very common today.
Or is the correct target cDNA preps for hematologist-oncologists to detect fusions in leukemia? Over multiple London Callings or Nanopore Community Meetings there have been compelling presentations on the value of moving these diagnostics out of large medical centers and into community settings. But is there a market?
Puckdx should be relatively flexible in these areas. The platform supports magnetic bead cleanups and thermocycling, so that ticks off two key capabilities to support most library prep workflows. But there would be a question of how many chambers are available in the device, the volumes available, how many chambers can be simultaneously incubated, etc. Another example might be cell-free DNA: Puckdx isn't going to handle the volumes of input involves, particularly from urine where it is import to process a large volume because the cfDNA concentration is very low.
Some the perpetual conundrum in the molecular diagnostics space: cool tech isn't enough. What Puckdx for sequencing strongly needs is a clinically-minded development partner willing to invest in a systematic commercialization, with some tolerance for exploring multiple applications until the most economically viable ones are discovered.

No comments:
Post a Comment