Tuesday, October 16, 2018

You Can Be Impatient Running MInIONs, But Not Feeding Them

Yes, it's been way too long since I wrote here.  Even longer since I did so with any regularity.  There was always some list of things draining my time and energy.  But I resolved this week to get back on the horse -- and that was even before today's bit of dilithium news. In particular, in one twenty-four hour span three different people remarked on the prolonged hiatus -- a professional contact, a commenter on the blog and finally some very cutting remarks from Draco (aka TNG).  And what better way to get going again but to kvetch about Oxford Nanopore's supply chain model?

We've been having some decent success with nanopore lately -- not top of the leaderboard success but useful productivity nonetheless -- but that has been marred by a run of flowcells that weren't good to start with or died in short order when stored in the fridge.  Oxford has been good about replacing them -- there's some minimal questioning about how you stored them and did you test them -- but the issue is the slow turnaround time.  It's like discovering your bananas have gone brown but the grocer won't get them for another week.  But first, a diversion.

We live in an age of rapid merchandise delivery.  I can order gadgets from Amazon and have them delivered to my door in a day or two -- or even have them delivered (except when Amazon screws up) to a locker in a remote location.  In the biotech world, I can leave purified DNA at the front desk by four in the afternoon and have Sanger traces by the next lunchtime.  Or I can design oligos this afternoon and have them by second breakfast.  

That sort of turnaround is common for what typical bench researchers need.  Ideas float into their heads and that leads to experimental designs which trigger purchasing desitions.   Obviously this isn't instant -- though there are freezer programs and such.  As an aside, I do wonder why NEB doesn't have a factory outlet in Kendall Square -- or at least a freezer in the Seven-Eleven in our complex.  Because, seriously, sometimes it turns out the ligase has gone bad -- or somebody took the last aliquote without notice.

Now big operations, core labs or big screening campaigns, may operate on less of a just-in-time basis.  That is a good idea, as a big operation doesn't want to stall for lack of parts.  We once had a near crisis when it turned out someone had taken their use-it-or-lose-it account at end of fiscal year and bought essentially the entire world supply of tips for our liquid handler; only the salespersons' private stash kept us from cooling our heels.   Or I've once had a vendor steer me to a different choice of Illumina sequencing choices due to a seemingly permanent backorder situation for one of the kits.  

Oxford really has feet in both worlds -- PromethIONs and GridIONs are clearly for the big operations, whereas MinIONs and Flongles and such are aimed at smaller researchers.  With its rapid turnaround and small footprint the MinION feels like something you should be able to launch whenever a good scientific idea comes to mind.  And in synthetic biology, that is a very real possibility -- I might dream up an experiment on the commute in, design primers during the day, order them and hope to get results the next day.

The catch with Nanopore is that flowcell and reagent ordering is nothing like an overnight affair.  NEB may not have a store in town, but I can order today and receive tomorrow.  But with ONT it appears from experience that shipments are basically once a week affair -- and there's no posted deadline for when to get your order in.  Which is another bit of poor design -- freaks like me get an incredible rush from slipping a HotPlate order in just before IDT's deadline!

So if you dream up your experiment on the wrong day -- or your flowcell tests out bad late in the week -- you may well be waiting well over a week to get a new one.  Which might give you plenty of time to design a less satisfactory but minimally sufficient alternative.  

A key decision Oxford made early on was to go it alone on distribution rather than partner with an established player in the biotech reagents and consumables market.  There's obvious attractions to that route -- great flexibility and control -- but the downside is one must learn lots of lessons and make many mistakes that an established vendor would avoid.  

Undoubtedly Oxford plans to someday have next-day delivery of consumables.  It's inconceivable that they wouldn't aspire to that.  But they aren't anywhere near that now -- and every day they stick to the batched delivery model runs a risk of further entrenching the procedures and mindset that separate such a daily delivery organization.

1 comment:

Anonymous said...

I work for a large(ish) sequencing company, and even though I think the Oxford nanopore technology is fairly decent and may become great if logistical problems can be sorted and they can deliver more regularly on promises. I see this company reviling big I and killing the Pacbio market for long reads. However, my colleagues don't even consider Oxford a viable competitor at all. Not sure if its just short-sightedness or we just confident in the development of other technologies. I cannot see how Illumina can get much better with the current chemistry as its quite mature at the moment and their improvements are small steps and not larger leaps like I see Oxford does. We also seem to think the treat is going to come from the East with BGI quite big in China, quite a huge market itself. anyways, as always I enjoy your updates on the progress of Oxford, please keep them coming.

P.S. Could you do a review of all current sequencing technologies available in 2018 and compared them. that would be cool.