{\rtf1\ansi\ansicpg1252\cocoartf1561\cocoasubrtf600
{\fonttbl\f0\fswiss\fcharset0 Helvetica;}
{\colortbl;\red255\green255\blue255;}
{\*\expandedcolortbl;;}
\paperw11900\paperh16840\margl1440\margr1440\vieww16180\viewh13200\viewkind0
\deftab720
\pard\pardeftab720\ri-340\partightenfactor0

\f0\fs24 \cf0 I'd be very surprised if you responded to all of these questions, but I've written out my thoughts and attached them. \
\
This is to give you an idea of what someone like me might think on reading your current slide desk, and the sort of questions I'd try to understand to get a better insight into your proposal, based on the material currently presented. I have a molecular biology background but not lots of instrumentation evaluation or PCR expertise, with a couple of years of low-level project management work. \
\
If any of them strike you as helpful/worth clarifying I encourage you to incorporate answers into your slides. Good luck with your work!\
\
\'97\'97\
Let\'92s assume I\'92m already persuaded that population testing is an essential good, and a \
goal that we should definitely be pursuing.\
\
Part 1: Method differences\
My understanding so far of how this method achieves that goal: the headline difference in using this method is that you are using endpoint PCR, not RT-qPCR. So, trying to explain that to myself, you look at the level of detectable viral DNA at the end rather than looking at a curve of viral DNA amplification and the point in time at which the cycle threshold (number of rounds of amplification) at which the amount of viral DNA exceeds a pre-defined threshold. So you lose the ability to do quantification and melting-curve analysis with endpoint PCR. The argument made is that it's not necessarily limiting in SARS-CoV-2 detection, because the information you want is not relating to quantitative information like viral load but instead you want a binary yes (lots of virus detected at one time point) / no (no virus detected at one time point).\
\
**So, the biggest assumption/assertion 1 is that an endpoint PCR assay is sufficient for diagnosis of SARS-CoV-2 vs. RT-qPCR.**\
\
I am not really qualified to assess this difference and wouldn't quite know what questions to ask to establish this. \
\
Part 2: Efficiency gains in processing\
Then there come the detail on efficiency gains: you're suggesting switching from manual handling to a fully automated decapping process. Presumably you'd still need tubes debagged in a class II microbiology safety cabinet, and as tube sizes vary you'd have to triage them for the type required in the decapping automation [they come in a range of different sizes]. I'm interested that the automated transfers work with swab in tube as that would very much help. There's no information I can see in your documentation about the technical details of how this would work. I can confirm this stage would be great to speed up and would love to see how this could work. At this stage I also imagine would be a need to triage barcoding and have a separate process for correcting badly coded tubes as there's no way they'll be applied to the tubes consistently.\
\
The second gain here is using 384 well plates rather than 96. This isn't specifically a limitation for RT-qPCR vs. endpoint PCR, but seems to be emerging from the availability of equipment for 384-wells for RT-qPCR process (e.g. Tecans, KingFishers, Tecans again, ABI7500s are all set up to work with 96-well plate formats). So the gain here would be that if you use a endpoint PCR process that you can provide sufficient volumes of equipment for that method that does handle 384+ samples per plate, and the consumables to match. But could there be 'halfway' improvement - e.g. if there was insistence on remaining with RT-qPCR, could the equipment be implemented for steps from decapping through to PCR reagent plating, then have it be loaded into real-time PCR equipment that can handle 384-well plates? Would that still provide at least some benefits? Or would there be losses - e.g. maximum volume limits - that would make this infeasible?\
\
Your documentation also mentions a applying different RNA extraction method - more information on that would be welcome too (currently as you probably know it's using MagMax kits).\
\
Then to the actual endpoint PCR: water-bath/thermocycler seems pretty straightforward, and I have no reason to assume this equipment wouldn't be capable of handling the temperature changes required. Presumably also the detection systems are well-established and validated technology.\
\
Part 3: How do we know all this could work?\
I guess I'd also be curious about how you'd plan to do the validation studies - presumably a scale-up from the existing sort of concordance data you have with many more sample plates? The current data is of course nice to see and I imagine quite difficult to obtain, but a 'next step' ask in terms of getting more concordance data would provide a clear direction for someone sufficiently well-resourced and motivated to evaluate your proposal further.\
\
It might be easier to understand the workflow if there was a detailed case study showing how it was already applied successfully to a diagnostic process (from the information in the deck it seems this is already being done at scale with some customers). \
\
Part 4: What limitations aren\'92t considered here?\
Of course, there's another limit on population sampling at the scale suggested here that is outside your control: sample collection in terms of swab administration and logistics. Presumably this procedure still uses nasopharyngeal swabs taken as the source material? We already have a lot of challenges surrounding taking these swabs from c.100,000 people, let alone 16 million pooled samples. Would be good to see this clearly spelled out in the slides and some consideration given to this challenge (even if just to point out that it will exist as a significant limitation that needs to be overcome with a lot of supply chain / logistics support).\
\
I imagine another limitation might be reagent availability - sure, it would be great to get the increased throughput with 384 wells and faster times from sample to result, but does the endpoint PCR method make the available reagents go further vs. the current RT-qPCR system and allow for scale-up even with that limitation? \
\
Part 5: Money\
On cost savings: the headline saving of \'a32.50 vs. \'a355 is obviously a big improvement. I know little about how the costing of this sort of thing works, but where is that saving coming from? (e.g. is it from less staff, cheaper reagents, cheaper equipment, reduced consumable bills, all of the above?) It might be easier to think about that reduction if it's mapped back to the concrete 'visualisable' places where the money is being saved.}