Remote design and execution of diy bio experiments

[Andrey Samokhvalov]

Hey, all!

I am thinking about bio cloud laboratory[1] targeted on the do-it-yourself bio community, a bit naive, but anyway I am curious what you think about it.
I would appreciate it if you share with me your concerns.

Also, I have prepared some questions, which might help me orient the idea in a proper direction. Pick one if you have a good mood :)

1. What is the biggest risk why it will not work from your point of view?
2. What is the biggest frustration/problem you have in DIY bio? Why it is a problem?
3. Is hands-on experience is essential for you, or you are more motivated by seeing the end result?
4. Have you been in a situation when you didn't have the required equipment/reagents and weren't able to buy it? Which one?
5. What were the first one/two experiments you did?
6. (optional) How much do you spend on DIY per year? Sorry for such a personal question, I am asking because such a project might require heavy investments, and I somehow need to calculate the market and stuff.

[1] Biology cloud laboratory - is an automated physical laboratory, which has access to it over the browser user interface, where you could design and execute the biological experiment remotely, and receive/see a resulting data over user interface or ask for resulting material to be shipped to your home. Examples include Emerald Cloud Lab, Transcriptic, and others, but as far as I am aware they are targeted on the pharmaceutical market. I tried to get access to their user interface, but they haven't replied : (

Thank you, and have a good day/night!

P.S. I am literally the first day here, so my apology if that was already discussed, or I violated some community rule.

[Dakota Hamill]

1.  Money.  You need a lot to buy equipment necessary to do the experiments, you also need to know how to run, upkeep, and service every instrument.  If you don't have redundant equipment, losing one piece stops your workflow.  Your user base by definition are DIYBio folks, which for the most part...want to do it themselves, or if not, 99% are just general population people, and I'd say it's safe to say, most of us aren't rich.

2. Science is hard.  It costs money (your own - someone else's if you're lucky) every time you run an experiment, whether it succeeds or fails, you still eat the cost.  Since most do DIYBio for pleasure, it's an expense, not an investment. Some are business/product minded though, so the ends justify the means.

3. Long past the point of wanting to know everything about everything.  Accepted you can be ok at a lot of things or really good at a few things.  It's 100% worth paying someone else to run experiments for you if they're better at it, and know how to get you want you want.   You get what you pay for.

4.  More times than you could count.  NMR, HPLC-MS, NanoDrop, UV/Vis, particular plasmids, particular restriction enzymes, MALDI-TOF, Sequencing, Microscope, laminar flow hood, bio-reactor, spray-dryer...this list could go on forever.  If there's a real need, you look on eBay, then dumpsters, then local lab-resellers, then ask local colleges. 

5. Isolating Penicillium from citrus fruit.  Isolating endophytic fungi from plant tissue.

6. This is hard, the lines between DIYBio work and daily science work is now blurred for me.  I never made more than $35k my first few years out of college at my job, but I'd say ~$5k a year on used equipment and reagents was typical.  Luckily being around Boston, you can find some very generous benefactors and supporters, for donated equipment, free samples from NEB for enzymes, etc.

As a note, being a CRO sucks, from what I've seen and experienced working at one, unless you're reaaaaaaaally good at what you do, or you're the only one that does it, or you came from Industry or Academia where you were trained at length on someone else's dime. 

People rent out time on their machines already, there are some websites that do that, it's a market you can list your equipment and an hourly fee or sample fee, with or without data analysis.

Not saying you have a bad idea, just hard to execute to make a profit.   Prove us all otherwise!

Think of your end customer...for the most part, an amateur, with little experience, little money, and not entirely sure what they want.  That's fine, we all start as amateurs.

What I'm saying is there's a reason Emerald Cloud caters to the pharmaceutical market, because they have a shitload of money. If they built Emerald Cloud (which is top tier equipment and talent) for a customer base of DIYBio people, they'd already be bankrupt, because that's a poor target market. 

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/4f0013c7-1db6-444f-b1ce-3b321671155c%40googlegroups.com.

[Peter Klipfel]

1a. Cost vs. Demand: There is not enough demand to make an automated lab profitable if it is serving DIY folks. You might be able to work if you cater to some specific part of the biotech space.

1b. Regulation. Making and distributing biological materials is fraught with regulatory issues. For example: what do you do if someone asks you to make anthrax?

2. Knowledge - the subject is deep, complex, and subtle. Understanding what can and cannot be done often takes me longer than doing an experiment. Such is research, I suppose.

3. I like working in the lab. Part of the reason biology is cool is because I get to move away from my computer.

4. I've had plenty of labs tell me that they can't ship to me because I am not affiliated with a lab. In these circumstances, I've had the local biohackerspace order it for me.

5. GFP transfection in e. coli. This is the standard "Hello world" of synthetic biology.

6. Maybe a few hundred dollars.

Andrey, feel free to reach out to me directly if you have any questions.

Peter

--

[Andrey Samokhvalov]

Thank you for such an extensive response!

-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diy...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en

Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.

To unsubscribe from this group and stop receiving emails from it, send an email to diy...@googlegroups.com.

[Andrey Samokhvalov]

Thank you, Peter, I will try to absorb both responses and come up with some follow up questions.

-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diy...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en

Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.

To unsubscribe from this group and stop receiving emails from it, send an email to diy...@googlegroups.com.

[Markos]

Hi Andrey,

Thank you very much for your message. :-)

I didn't know the work of those companies that you mentioned.

Just a suggestion. Have you considered offering this type of service to schools?

Students could conduct educational experiments in chemistry, physics and biology in remote laboratories.

And the school would pay for access to the service.

It would not be the a virtual environment of a simulator.

But a real environment, subject to all variables (and complexity) of the real world, but only with remote access.

Best Regards,

Markos

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en

Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.

To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.

[John Griessen]

On 6/2/20 7:22 AM, Markos wrote:
> Students could conduct educational experiments in chemistry, physics and biology in remote laboratories.
>
> And the school would pay for access to the service.
>
> It would not be the a virtual environment of a simulator.
>
> But a real environment, subject to all variables (and complexity) of the real world, but only with remote access.
>

Wouldn't it be nice to have a camera view of *YOUR* tubes and vials as they moved through the robot handler system to gather
reagents, have processes like PCR done, land in their racks waiting for next steps? That might get young students more interested.

[Fork Face]

That would be nice. My class is a hot mess because everyone’s yearlong project got nixed because of covid. Everyone’s trying to put something together “virtually” right now and it’s not great.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/26bb5416-1ef6-09b6-80e2-e8d7c684fd89%40c2o.pro.br.

 

[Jonathan Cline]

I performed remote synbio experimentation over the cloud many years ago now.  (Seems like a lifetime ago.)    I had two webcams and an industry biorobot and I ran unix-style commands on the laboratory PC using cygwin for network cloud operations and biorobot commands.  I wrote both ends of the software for biorobot control; the server side to operate in the lab, and the client side to issue commands remotely (my Perl Robotics software).   At the home office I ran VideoLAN to receive the two video streams.  

The remote-operation results would be much more hi res today, with 1080p cameras, compared to the 360p webcams I was using.  Frame rate was surprisingly good.  Network latency (even up to 500 ms or so) was surprisingly not much of a problem.  The experiments I ran were for DNA preprep and successful with caveats.  Biotech robotic platforms have extreme downsides.  They do not handle mechanical or liquid tolerance faults well. I constantly had to run flush steps due to robot handler error (time consuming and expensive).  Laboratory access is always required for specific manual labor such as reloading new tips, refilling reagent reservoirs, and sometimes issuing a hard equipment reboot.  The equipment reboots would be less of a problem these days as the embedded platforms have become much more stable.  DO NOT USE MICROSOFT WINDOWS. 

I was able to run complete protocols, start to finish.  The most successful completion was with the version using magbeads which are very forgiving with biorobotics but expensive and volatile.   The physical platform needed to be carefully configured and calibrated for the protocol.  It is not good for arbitrary protocols or 1-off projects.  That is the major downside.

Based on the expense, complexity, and large initial R&D needed per protocol, I don't believe it is a good fit for an ad hoc DIYbio style environment.  It is very useful for highly repetitive experimentation or high volume manufacturing in a controlled environment.

I don't think "browser based" would be on the desirable list of requirements.  Use the command-line and text-data files.  It is more powerful, more consistent, and scriptable, especially for a batch operation.  A browser interface would be significant over-design and introduce more problematic layers.  "User interface" is the very least of the concerns.  Consider the main problem to be the reliably of sucking up a microliter of liquid and depositing a single, specifically-sized droplet into a very tiny tube, as well as the repeatability of this operation.  Having a fancy javascript UI frontend to this ability is adding risk to each experimental run.   Consider the more real problems to be precise humidity and temperature control of the lab room, keeping the camera lenses clean, avoiding arm collisions due to equipment or protocol error, etc.

On Friday, May 29, 2020 at 11:03:36 PM UTC-7, Andrey Samokhvalov wrote:

Hey, all!

I am thinking about bio cloud laboratory[1] targeted on the do-it-yourself bio community, a bit naive, but anyway I am curious what you think about it.
I would appreciate it if you share with me your concerns.

 

-- 

## Jonathan Cline

## jcl...@ieee.org

## Mobile: +1-805-617-0223

########################

[John Griessen]

On 6/3/20 1:05 AM, Jonathan Cline wrote:
Consider the main problem

> to be the reliability of sucking up a microliter of liquid and depositing a single, specifically-sized droplet into a very tiny tube,

> as well as the repeatability of this operation.

> Consider the more real problems to be precise humidity and temperature control of the lab room, keeping the camera lenses clean,
> avoiding arm collisions due to equipment or protocol error, etc.

Thanks for these reality checks. Do you think the error and repeatability of pipettors is from
the air over liquid methods, or something else? I've seen some positive-displacement-syringes that use no air, just a metal wire
plunger in a close fitting glass tube. Is that something that needs more testing for automated liquid handling success? Or is
there a need for a better automatable seal between pipettor tip and pipettor -- maybe with an O-ring?

[John Griessen]

> On 6/3/20 1:05 AM, Jonathan Cline wrote:
> Consider the main problem
>> to be the reliability of sucking up a microliter of liquid and depositing a single, specifically-sized droplet into a very tiny
>> tube, as well as the repeatability of this operation.
>
> > Consider the more real problems to be precise humidity and temperature control of the lab room

Just a change in temperature while dispensing changes the cal of an air over liquid pipettor, doesn't it?

[Jonathan Cline]

I'll say this:

If MIT Engineering departments spent any time on biorobotics design
competitions rather than sponsoring publicity-happy robot soccer
tournaments, the field of synbio would be much farther along.

On 6/3/20, John Griessen <jo...@industromatic.com> wrote:

>
> I've seen some
> positive-displacement-syringes that use no air, just a metal wire
> plunger in a close fitting glass tube. Is that something that needs more
> testing for automated liquid handling success? Or is
> there a need for a better automatable seal between pipettor tip and pipettor
> -- maybe with an O-ring?

The Tecan platforms use water displacement, not air, if I get your
meaning. Humidity differences cause changes in evaporation. There is
no problem of the seal of the tip cartridge, there is no leakage, the
plastic tips are form fitting. The chasm from my perspective is that
a human biologist will make numerous visual error-checks while
pipetting, to ensure a droplet is precise, this involves many things,
such as holding a tip at an angle to the target, tapping with a
fingernail, visually checking again, dispensing less or more based on
droplet size, etc. A robot arm can't do these things. Also note that
viscosity of the various liquids is variable. Using magbeads modified
this problem because they are in suspension. Many of these types of
problems are the reason research has been put into alternate carrier
methods not requiring traditional liquid pipetting, such as
oil-in-water microfluidic experiments but those weren't very viable
either (haven't checked progress lately).

A random whitepaper which might or might not be interesting reading.
https://www.pacb.com/wp-content/uploads/2015/09/Guide-Pacific-Biosciences-Template-Preparation-and-Sequencing.pdf

Overall the point being that biologists don't enjoy automation for a
simple reason: it's annoyingly difficult and time consuming therefore
not worth investing in, unless specifically used for very high volume
experimentation. Since DIYbio is more ad-hoc 1-off experimentation,
it does not really make sense to apply the technology.

[Andrey Samokhvalov]

Emerald Cloud Lab agreed to show me a 30 minutes demo of there UI and lab, but probably I will be able to ask some additional questions.

Q: What questions would you be interested to ask them?

Write questions here, and probably I will be able to forward them.