24 replies to this topic

[manofsan]

I really enjoy reading CNN's Business2.0 site, because they always talk about the latest business trends and innovations.

Here they have an article on how ordinary basement startups may be poised to bring fresh innovation to the biotech industry:

http://money.cnn.com...otech/index.htm

But here's a thought to quicken your pulse: Big changes are afoot in the life sciences, and basement biotech is neither as far away nor as expensive as you might think. In fact, you can buy all the equipment you need - DNA synthesizers, sequencers, and so on - for as little as $50,000 on eBay

I'm rooting for these people, because we need to see more competition in that marketplace. I'm particularly rooting for those who'll provide the fundamental enabling products that unleash the creativity of others -- the bioinformatics tools, the lab-on-a-chip devices, the gene-parsing and analysis software, the low-cost rapid gene-sequencers, the gene synthesizers, the gene-parsing and analysis software, cheaper and smaller protein crystallography systems, etc, etc.

But in which area will we first see movement coming from? Which part of the logjam or gordian knot will loosen up first? I'm thinking it's the lab-on-a-chip devices which will create the most movement in the biotech industry.

If you were given the opportunity to start up your own home biotech shop, how would you go about it?
Which market would you pursue?

I'm thinking that industrial waste processing or crop genomics might be the easiest, providing the quickest Return on Investment.

How about looking at some biochemical process for which existing natural bacteria are employed, and then finding some way to enhance them to do the job better? Any particularly standout candidates?

Edited by manofsan, 17 June 2006 - 04:57 PM.

[Live Forever]

I always root for the little guy anyway, but this seems to be especially good.

[maestro949]

If you were given the opportunity to start up your own home biotech shop, how would you go about it?
Which market would you pursue?

I'm still doing the due diligence myself. I'm looking for an aging-related bioinformatics project to dive into but haven't found anything yet except perhaps building a database of aging related genes, proteins, pathways, etc. Data accumulation doens't seem that valuable with the number of cross DB search tools that are evolving. The one thing I'm interested in is pathway simulation but that's way beyond my feeble skills ATM. In the meantime, while I look, I study the science. There's no shortage of stuff to learn regardless what level you're at! [mellow]

[manofsan]

But remember, there are all kinds of useful projects that might not be directly and explicitly be equated with anti-aging, but which might otherwise have a beneficial impact on bodily health. The tooth decay thing I just posted would be an example. It's not explicitly elevating your lifespan, but tooth decay is a symptom often afflicting the elderly, and it does impact your quality of life.

Moreover, it might be a stepping stone into research that is indeed more directly connected to lengthening lifespan. There may indeed be a number of anti-aging precursor markets that can be pursued via biotech. Agricultural genomics might be one of them.

[John Schloendorn]

When fighting aging is what you want, then doing something that is not fighting aging is the most ridiculous and most disastrous mistake one can possibly make, imo.

[manofsan]

But John, the tools used for fighting aging may in fact be the same as the ones used for stopping tooth decay, acne, hair loss, obesity, etc. (ie. these cosmetic problems may involve the same genetic engineering techniques, and they may have an immediately wider market appeal)

Actually, consider that losing an arm or leg or an eye may not have a direct impact on your longevity, but I still support stem cell research, because regenerative medicine may also help longevity, and not just growing back an eye, leg, or arm.

[John Schloendorn]

The gradient of my agreement with your statements goes stem cells > obesity > tooth decay > agricultural genomics (total agreement > total disagreement in that order ;-)

[lunarsolarpower]

If you were given the opportunity to start up your own home biotech shop, how would you go about it?
Which market would you pursue?

The most frustrating thing that struck me during my short stint as a molecular biology researcher was the vast amount of repetitive labor that needs to be performed by human lab techs. Although many of the processes are routine, a human must physically do each step by hand. I would love to see a device invented that would resemble a very large microwave. The scientist would design the experiment and determine the precise details (number of replicates, course to follow depending upon results, etc.), push the button and walk away. 3-30 days later perfectly identical results would roll out whether that be a western blot, a spectrophotometry reading, a GC Mass Spec chart, or the sequence of a gene. While no one box could perform all the science that goes on in a modern lab, perhaps they could be designed to be modules that could be chained together to collaborate on various processes. There could be a cell culture module that would combine the sterile hood and the incubator in one, an instrument module that would have access to digital data collection devices, etc.

For the large scale techniques that represent the near future of molecular biology (directed evolution for one) massively automated parallel experimental setups are crucial. Has anyone else ever thought that something like this would be of value?

[maestro949]

Yup. Anything you can automate (or sidestep) in this process is worth a pretty penny to big pharma and will lend itself well to the fight on aging when it's our turn at bat. Doing it in your garage? Hmmm...




Source: http://interpro.engi...pharmaceutical/

[manofsan]

Nice comments, guys. Well, there's no reason why that flowchart process couldn't be automated and organized by a software package. Laboratory Information Management Systems (LIMS) were invented decades ago, and have only improved since then.

There are umpteen varieties of open-source Linux, I wonder if anyone has one that's suited for an open-source LIMS? If not, then it could become a promising project that could attract people from the open-source community to develop something.

People are also talking about Ruby-on-Rails as a marvelously easy platform for creating interactive data input and retrieval applications. Maybe something like that could used to create a generic universal one-size-fits-all turnkey solution for LIMS purposes.


The reason why I mention those lab-on-a-chip devices a lot, because they are so much like the integrated circuits which replaced vacuum tubes -- yet another case of making glassware disappear. Doesn't anyone remember those Radio Shack home electronics kits, or where you could buy an entire science lab in a box? Efstone? Why can't technology reach the point where you can buy a self-contained gene-analysis kit just to teach some science in a highschool classroom?

Either way, automated lab machinery would simply become peripherals requiring device drivers, just like the printers and scanners we all use.

I'll particularly note the role of companies like Hewlett-Packard which built up the PC printer industry into the sophistication it has today, with fundamental and pioneering innovations in electromechanical systems. One of their more advanced divisions was spun off into Agilent.

We need an HP to span the gap between the information-rich world of biomolecules and the information-manipulable world of computers.

Even IBM has a reputation for being notoriously far-sighted in releasing protected technologies into the public domain to seed and spur greate development of future markets.

Perhaps we need to hunt down more people in the relevant fields while online, in order to obtain their more direct insight on where the most hindering lab-technical-process hurdles are.

[manofsan]

While Intel Corporation may be a classical example of a heavily-capitalized enterprise founded by PhD's who'd made fundamental technological breakthroughs, Apple Computer is a classical example of the garage startup that helped to kick off a multi-billion dollar industry.

Perhaps the following are frivolous and rudimentary examples, but they may give some ideas on the direction that things should go in to make lab automation more available to the masses, and for basement startup laboratories to take off:

http://linuxdevices....8960820667.html

http://www.vnunet.co...open-source-toy

http://oap.sourceforge.net/

Maybe we need to invite some of these automation-Gutenbergs over here, to give their perspectives and discuss.
Building a toy robot to navigate your hallway is cute. But it might be a whole lot more inspiring and motivating to innovate do-it-yourself lab automation that gives rise to a new generation of startup basement labs that might make vital headway in solving the aging problem. The massively parallel approach?

Please read this:

http://www.bioforge....a?messageID=266

http://www.bioforge....ID=336&tstart=0

Comments?

Edited by manofsan, 20 June 2006 - 01:42 AM.

[maestro949]

I like the way you think.

Biotech is chewing through generations of biomedical technology like mad. One idea for an open source or angel funded team to tackle an aging related project would be to snarf up some of this equipment on the 2nd hand market, i.e. Ebay. The companies that produce this equipment usually post the user manuals and support FAQs on their websites. A lot of bioinformatics software has gone open source already. I've been downloading quite a bit of it and tinkering with it. The quality of many are quite good. Many though are solo efforts though that fail to build momentum.

I see a nice intersection of used biotech equipment, open databases of genetic and proteomic data, open source software and open science journals all meshing for a collaborative anti-aging research project on the horizon for ImmInst.org. We should watch The Synaptic Leap(link to article), bioforge, etc. and follow in the trail that they blaze. The biggest challenge will be resourcing such an effort with the appropriate talent and leadership. Perhaps an RFP is in order to solicit some angel funding to even research the possibility of such a project.

[manofsan]

Microsoft Robotics Studio has just been announced:

http://blogs.msdn.co...io/default.aspx

Could this be suitably adapted for laboratory automation purposes?

[maestro949]

Interesting. Microsoft has also added Physics to it's DirectX API and is putting up money for people to run NT clusters for science projects. (it might be worth taking the money and then switching back to Linux after a while). Bill Gates just announce his retirement. I can only presume he's coming here to help us with the aging fight.

[manofsan]

LoL, physics inside DirectX is only for videogaming purposes, so that videogamers can watch boulders tumble, etc more realistically by using dedicated physics hardware instead of straining their CPU.

But if Microsoft is trying to streamline the robotics pipeline with its own robotics API, then perhaps it will positively benefit lab automation efforts.

I still think lab-on-a-chip devices will provide the most efficient and streamlined processing of physical samples.

You saw that Business2.0 article refer to companies like Cadence Design Systems, which produces software used by computer-chip designers to produce new chip designs and architectures. What they were saying was that if there were similar design software for lab-on-a-chip devices, then this would result in more small design houses forming in order to create new custom designs for lab-on-a-chip systems. In the computer chip industry there are already plenty of small chip design houses, called "fabless chipmakers" because they just design, and outsource the actual chip manufacturing to those who do have chip-making factories.

But imho, a lot of the money and motivation to do this would depend on the connection with an end-market. Right now, the end-consumers for these products would just be labs, since molecular biology and genomics can't simply churn out products direct for the masses, like with iPods and PCs. So if labs are the end-market for lab-on-a-chip products, then competition and evolution among those products will have to grow with the lab market.

I wonder if leasing will become a more viable option for high-end analytical equipment, as compared to purchasing second-hand, etc. Let's compare with the photocopier industry for a moment. Most modern photocopiers that you see in the workplace are fairly sophisticated and expensive devices, although they're pretty commonplace. Many of them are even leased rather than purchased outright. This enables leasers to take advantage of equipment upgrades periodically as well. Could the marketing model for automated lab analytical equipment then grow to resemble that of photocopiers?

[manofsan]

A little more on the Microsoft Robotics Studio:

http://www.techrevie...7016&ch=biztech

a programming environment for writing and debugging software for robots that's similar to Visual Studio, the company's main tool for writing Windows software; a "runtime" environment that functions as a mini-operating system for robots, executing the code people write using the programming tool; and a simulator that allows users to build virtual models of robots and test how their software would behave on them, without having to build actual hardware.

So I liked that little mention of a simulator, since it could allow simulation of automated processes which could then be used as a way to design and roughly test out what your automation code would do. So then you could then create an API library for fundamental processes of biochemical analysis, which you could then use as building blocks to construct automated experimental procedures.

So from a lab automation perspective, this could be kind of interesting. It would still be nice if this could work with lab-on-a-chip devices, though.

Trower says Robotics Studio is intended to help the robot industry "bootstrap itself," the same way Microsoft's first DOS operating system provided a standard platform that other software writers were then able to use to write a host of applications, such as spreadsheets and word-processing programs, that eventually made PCs indispensable.

Well, anything that accelerates this bootstrapping process is a good thing.

[maestro949]

LoL, physics inside DirectX is only for videogaming purposes, so that videogamers can watch boulders tumble, etc more realistically by using dedicated physics hardware instead of straining their CPU.

It'll be interesting to see. If it can abstract a boulder falling on an orc. Why not use the same physics hardware for viewing molecules bouncing around?

So then you could then create an API library for fundamental processes of biochemical analysis, which you could then use as building blocks to construct automated experimental procedures.

But are these robots going to be able to work at the molecular level? I would think the lab-on-a-chip idea would involve nanotechnology combined with both an automation API and a cheminformatics API.

I wonder if leasing will become a more viable option for high-end analytical equipment, as compared to purchasing second-hand, etc.

Considering that the larger of these devices may be idle most of the time, collaborative efforts to build and share them somehow would be my suggestion. Think supercomputers and super colliders. If you came up with a worthy experiment for either, you could probably find time on one. The same could probably be true for the in silico labs of the future.

[opales]

Do it yourself DNA hacks:

http://www.dnahack.com/index.html

[manofsan]

Hmm, since we've been talking about the Methionine Restriction diet lately, this means we could try out diet-induced metabolic hacks too. (Standard disclaimers: always consult your doctor, blahblahblah)

[manofsan]

Here's an article in Wired on homebrew biotech:

http://www.wired.com...ml?tw=rss.index

and another warning about security implications:

http://www.techrevie...7122&ch=biotech

[maestro949]

Good resources opales and manosam.

From the DNA hacks website that you posted a link to..

Interesting thought, from Michael Schrage in the June 2003 Technology Review: "Maybe bathtub biotech will be the next to capture the mindshare of the techie tinkerers. Maybe bioinformatics and the diffusion of genetic engineering technologies will inspire a new generation of bio-hackers. Certainly the technologies are there for those inclinded to genetically edit their plants or pets. Maybe a mouse or E. coli genome becomes the next operating system for hobbyists to profitably twiddle. Perhaps this decade will bring a Linus Torvalds or Bill Gates of bio-hackerdom -- a hobbyist-turned-entrepreneur who can simultaneously innovate and market his or her DNA-driven ideas."

[manofsan]

hi maestro, that's why I'd previously suggested plant genomics, because it's an easier model to tinker on. Once you've developed a higher level of proficiency, then you could go on to the more complex systems.

Has anybody heard of http://www.ActiveCollab.com and also http://www.php-collab.com ?

They are free project collaboration service sites, enabling people spread out across a variety of locations to collaborate jointly on a project. Through their websites, you can track and manage your progress.

For a more detailed description, go to http://www.basecamphq.com which is for Basecamp, a pay site offering a similar service. The ones I just previously mentioned are free.

I think that online collaboration tools could also help disparate groups of people combine and amplify each others' efforts.

[zerowave]

Most universities allow outside users access to shared campus facilities, for very small fees. The best instruments cost a lot. Why make your own? Just rent labspace and instrument time.

I do think it highly likely that a 'basement' biotech startup could achieve a strong degree of success. Current research papers are published and available. Researchers are usually very friendly and love to talk about their work. There is plenty of knowledge out there and many opportunities for unique and lucritive contribution.

Edited by zerowave, 07 July 2006 - 11:18 PM.

[icyT]

These look really cool, I'm going to sign up for em all now so that when I finally do this I'll have contacts set up to start!

As for BaseCamp, there are paying features, but you can still use it for free if you want.

You're just limited to managing one project at a time (but hey, you can still participate in other ones, and how much do you want to manage at once anyway?)

You can't share files. You still have unlimited people and clients like the other ones, and you can use the Chatrooms, and you get 2 writeboards instead of unlimited like the paying people have.

Don't even need a credit card! Hell I'm signing...

[maestro949]

hi maestro, that's why I'd previously suggested plant genomics, because it's an easier model to tinker on. Once you've developed a higher level of proficiency, then you could go on to the more complex systems.

Yeah, personally though I hope to do my work in-silico with software based on the data pouring out of ongoing and future research efforts. There are plenty of reductionists tinkering with organic material and collecting a glut of data but in my opinion there are too few building and using models, simulation and analytical tools like GROMACs to do the necessary multivariate analysis that will be required for major leaps forward most of us here hope for. In my limited research I've convinced myself that even if someone gave us a blueprint of every detail of the proteome (functional and physiological) it would take us many many years how to figure out how to make even small adjustments that increase lifespan. I don't believe in a nearterm singularity nor that strong AI will be our savior. My gut tells me it's going to require brute force in-silico trial and error as doing this on organic life's timescale isn't going to cut it. The next-gen tools for doing this brute force work are evolving today and will trickle down to us mere enthusiasts in time for us to use in our virtual garages. Some clever people will take these tools and will ingeniously combine them with other concepts and catapult us forward an enormous order of magnitude. That, I believe is our only hope.

I think that online collaboration tools could also help disparate groups of people combine and amplify each others' efforts.

Indeed. It's quite easy to setup a virtual shop on the internet. The bigger challenges are time, vision, leadership, organization, human nature, funding and oh yeah, a working knowledge of the domain .

I do think it highly likely that a 'basement' biotech startup could achieve a strong degree of success.

The only measure of success for me is whether I can find a way to contribute to significantly extending my and many others' lifespan. Successfully navigating a startup to the point of aquisition or onto the fortune 500 list, acquiring gobs of wealth and dying at 90 is pointless.

Current research papers are published and available. Researchers are usually very friendly and love to talk about their work. There is plenty of knowledge out there and many opportunities for unique and lucritive contribution.

Unique? Yes. Lucrative? With the right combination of luck, skill and timing perhaps but the barriers to entry can be high. A good chunk of VC money is usually needed these days to hire the talent to even demonstrate whether you have a potential concept.

Edited by maestro949, 07 August 2006 - 10:13 AM.