Datomic .. not just because of datalog, but because its hands down the best implementation of a AWS lambda based workflow I've seen (Datomic Ions). It's such a peach to work with.
We're a nanopore sequencing shop. Couldn't agree with you more that nanopore generally is the future, not just for sequencing. Can't wait for ONT's solid state nanopore flow cells - you may get flowcells that go 2/3 times as long then.
But then for most things .. the problem is the prep, not just because the whole portability thing goes for a toss. Yay, great the sequencer is the size of a USB drive, but the rest of the lab isn't :/
More worryingly the biggest and most stubborn cost is now the prep, not the sequencing.
As you correctly pointed out - need to find a way around amplification - then both of these problems above go away if you can do direct PCR free sequencing.
The other less mentioned problem related to the above is also the need for parallelisation - some of those ultra low costs you read about can only be realised when you sufficiently multiplex your samples. For instance, its about 100 USD per reaction for ligation (last step of prep before the sequencing starts), you generally wait till you're sequencing atleast 12 samples in the same reaction so that you're paying <10$ per sample, not 100$ per sample, which is obviously insane.
I'm pretty sure they will come out with it. The protein nanopores were the first wave of nanopore research - its tried tested and stable now so they stick with it.
A few years after they launched, the first solid state nanopores were being demonstrated in the lab. Commercialising solid state nanopores seems to be easier if anything than protein nanopores because they slot right into silicon fabrication.
On PCR/barcoding ..
Yeah, thats right - do it right in the PCR step .. sometimes we avoid it, if we are not yet sure about the protocol. I think what I meant to say is that the full promise of nanopore sequencing for me is achieved only when you can skip having to amplify/multiplex/barcode - just extract dna, wash, add sequencing adaptors and go - for almost anything.
I think the way they are talked about people generally come in expecting that TODAY .. they think they can literally stick a single sample with no prep and get 1gb of sequencing done for 10$ in an hour. I've seen quite a lot of that. (even from people with PhD's :) )
So yeah its more that the minute you go PCR, you're in for a minimum 20$ per sample, often its the highest cost line item in your whole process.
If you're doing things like 16S metagenomics, you get sequencing at 2-3$ and prep at 10 times that, starts to feel "wrong" after a while, if you get what I mean.
Why're trying everything we can to make sure we're running at full capacity so that we can always give low prices even if its single samples.
Also seeing if we can reduce prep cost with microfluidics/mems - they have voltrax for this, but there are a few other vendors in the market. That has the positive knock-on effect of also reducing labour cost.
We're working on this, its fun, and already possible at lab scale. Check out Surface Enhanced Raman Spectroscopy (SERS).
And we've got a huge shot in the arm with covid. At consumer scale the hard problem is skipping any kind of sample prep. Requires unrealistically high sensitivity - we're trying to work around this with different no-prep based amplification techniques. There is actually a lot of interesting work happening in this space.
Further, currently the substrates use gold to enhance the raman scattering of the incident light. As you can imagine, this can get pretty expensive. There has been some success in using low cost paper/inkjet, textile/dye based approaches to introduce gold nanoparticles onto more familiar and easily mass produced substrate to achieve the same effect.
SERS is pretty cool, hasn't got its due and I think its going to be one of the underdogs in the diagnostic/detection market in the coming years.
Whats especially cool about it is - it doesn't restrict you to one thing - only genomes, only proteins etc.
You may need a different strategy for different pathogens, detecting the a gene or cell wall for one, a synthesised protein for another, or even the waste product / metabolite to check if a reaction took place - its not a hammer looking for a nail.
Once a month, I find myself wishing Clojure had static typing, before dramatically rediscovering why not having it can be such a relief. The reasons for this are expressed much more articulately than I can here: https://lispcast.com/clojure-and-types/ ... Personally, for the kind of work that I do, I've found clojure's compromises make my life on net easier, and my code, on net, more reasonable.
I'm finding that it really depends heavily on the domain I'm working in. Roughly speaking, if I'm dealing with "business problems" type stuff - grab data, report on it, keep a paper trail, etc. - I want a very loose and flexible programming environment with a REPL and dynamic typing and all that. Python, for example, treats me very well when I'm wearing my "data scientist" hat.
OTOH, when I'm doing more "building infrastructure" type work - implementing a data store, writing a compiler or interpreter, stuff like that - I start getting more interested in rigidity and formality. Static languages treat me well in these situations. The more rigid, the better - I'll prefer Scala to Java, for example, specifically because it gives me more tools in the type safety department.
I pretty much agree with this. When I can get away with it performance-wise (and I almost always can for personal projects), this is a big part of why I love using Perl 6.
I recently rolled my own (verg tiny and specific) build/deployment system for a project. Initial prototyping was a breeze, but once I had things more or less nailed down, I started adding types and, for example, was able to leverage the type system to make sure it can only be deployed listening on IP addresses in the private range. The next step is to rig up some sort of Zerotier integration to ensure I only make it available on my private networks.
Yes, I wish I had clarified that I meant that Clojure beats everything else for me for what you describe as "business problems" and also, I've found it quite enjoyable for writing simple games. Im afraid im not well qualified to comment on what "infrastucture" type work would be like, and what is best for it.
