1. DIY for "$0" is fine but 98% of those even on this list just want to buy a ready-made product. Myself included. There are reliability and reproducibility and other -ity concerns if self-building custom equipment even if it is to a common spec. It's mechanical engineering, not software. The amount of time and effort to prep and run simple bio experiments can't be understated so having to build machines first is a no go. Not even Woz sold many Apple I kits compared to how many ready-made computers he sold. Can't standardize protocols and parts if everyone is using one-off custom builds.
Besides which, most DIY hardware designs are missing this critical aspect to them called: *high priced engineering talent* which optimizes the design in part driven by cost constraints (capitalism works, ya know).
2. Cheaper reagents are being done and by some on this list but slow going. And patent minefield is worse there. DIY "grow your own" reagents are popular in research labs but again same concerns as #1, reliability and reproducibility and other -ity, when you buy a mix you are buying (supposedly!) the Q/A process not just the liquid itself. Reliability in part is a weakness due to longevity (reagents expire and/or are costly to maintain in a freezer), imagine if DIY computer hackers had to constantly re-generate their own silicon wafers to keep building new circuits.. And don't want people keeping reagents in their kitchen freezer next to their frozen vegetables and coconut ice cream (side note; sane biologists would not buy dairy ice cream).
3. Yes and Yes. Sample density is still a reason to go microfluidic. 96 wells is okay for humans, 384 wells is hard to deal with by hand, 1536 wells is ridiculous. The curve of available technology for data crunching means experiments should be run in "massively parallel numbers" yet one big limitation of this is simply the logistics of dealing with 1,536 separate experiments, whether handling them or tracking them. As well as the desire to track the experimental effects on a single cell rather than a large culture of diverse and perhaps-behaving differently cells, that means, the environment and sensors have to be similar scale to the cell.
## Jonathan Cline ## jcline@ieee.org ## Mobile: +1-805-617-0223 ########################On 1/28/16 8:13 AM, Simon Quellen Field wrote:
- Patent fences should not bother DIY non-profit experimenters. Let them try to sue me for my $0 profit.
- We have been discussing cheaper PCR machines. Perhaps we should be looking into cheaper consumables. At two million dollars a liter, there would seem to be some headroom on price. How about engineering an organism that produces the expensive part, and engineering a low-cost way of isolating it.
- Is the expense of the consumables the main reason for going microfluidic? If larger amounts are easier to work with, and the consumables were cheap, would people bother with tiny volumes?
On Thu, Jan 28, 2016 at 5:33 AM, BraveScience <bravescience@gmail.com> wrote:
Hi all,
@Ujjwal: that's a cool question. Personally, as it has pointed out above, connectivity is an issue. Even benchtop lab thermocyclers are kind of a pain to control smoothly. Actually they are horrible machines. Although I am part of the professional clients and my needs aren't always the DIYbioer's needs..Price as well is challenging. There's a lot of technology to assemble and run properly and reliably. And price is very important for professional too.
All the suggestions so far pointed on making the system more efficient/cheaper and increased user experience (my point as well).It's kind of remarkable, as well as happened for the fax, how thermocyclers didn't really change. They were horrible boxes and they still are.Like we weren't creative enough to come up with other systems that allow to heat/cool down some small amount of samples.
A good point was made by Jonathan: it's not too important to make a "cheaper" PCR machine by itself, but it's mandatory to reduce the amount of sample volume. What you can do in 50-100uL today you could easily scale down to 1-10uL. If you scale down to 1uL mix, well, it would be much much cheaper.During colony PCR we usually use 10-20uL total volume. That means 5-10uL myTaq mix (all included). And it always works. I even used it for cloning, fuck those proofreading crappy polymerases.
let's crunch some numbers: price tag is 104$ for 50uL reactions. That means 25uL master mix. Each reaction is 2$.Going down by 10x means 5uL reactions would mean reaching 0.20$/reaction. Even in today's lab you can do quite a lot with 5 uL PCR volume if you reach a final yield of 100ng/uL. Enough to do plenty of cloning IF you avoid those horrible spin column/gel extraction kits... Gel extraction looses up to 90% of your stuff, especially if it >5kb.
Anyhow my point is size. Yet, and i know many will share this point, working with less than 10uL is a pain.If you work in a tube.Because pipetting isn't easy after the psychological barrier of <2uL and capillarity is a bitch.
Why don't we leave behind our backs those tubes? Come up with something new.Kind of muhammad story for molecular biology, if you cannot bring the heat to the sample, bring the sample to the heat.Samples could move across temperature gradients. You could get either low volume sizes and user friendly equipment.Illumina and nanopore are already doing this. All sample preparations is ran on digital microfluidics biochips, although a patent fence as high as they sky is out there, and they perform an enriching amplification step previous to sequencing.
Best,Fede
0 comments:
Post a Comment