Intern from Teselagen here. <shameless plug> We have a private Beta that you can sign up for on our website if editing your own DNA sequences is appealing to you: http://www.teselagen.com/user/registration </shameless plug>
It's pretty cool to see Synthetic Biology getting coverage on HN. One of the really interesting ideas that scientists are starting to grab a hold of is that of "biological parts". People have started compiling these biological circuit elements and figuring out how they work when exposed to different stimuli [1]. The ultimate aim is to standardize the usage of these parts so that you could design an organism in your browser, figure out whether it will work or not (based on a super-secret algorithm), and get the resulting functional organism shipped to your lab bench in a couple of days. As a reference, right now it can take up to 2 months to even create an initial DNA product with traditional methods. From there, it could take months to get your organism working correctly. New DNA synthesis techniques will dramatically speed up some of the initial steps (something Teselagen has done).
An addition to the article I'd make is that Teselagen is actually letting people leverage new DNA synthesis techniques produced by a couple of groups (the Venter institute is one) to design better biological circuits (as in faster, better, stronger). If you are interested in reading about how scientists have begun to overcome some of the many barriers to biological circuit design, there's a good intro to advances in DNA synthesis techniques here: http://j5.jbei.org/j5manual/pages/1.html
Co-founder at TeselaGen here. Excited to see more discussion about Synthetic Biology in HN since we are all hackers. I'm also happy to see an excited intern spontaneously talking about our company :) Feel free to register, however expect some delay (hopefully not too long) before you get an invitation to test our new UI. Also, here is a cool video that you guys might like:
http://www.youtube.com/watch?v=3xArETjua30
We haven't talked too much about what we are doing but if you are one of us, who share the vision of moving Synthetic Biology to the cloud, and are excited about the development of cool technologies to automate the designing, building and testing of novel strains, contact us. We are developing a lot of software and will begin with software/hardware/wetware integration soon.
My apologies frisco. I'm co-founder at TeselaGen. We've been running like crazy, currently engaging with some of the largest biotech companies in the world. We would love to have you as a user and get your input as well. We've been holding invitations until we release our upcoming new version, later this summer. We'll be contacting you soon!
> DNA needs to be read. If you want to identify the genetic code of something you need to sequence it. This is like being able to open the source code other people's programs to see how things work, debug your experiment and is the analogue of an oscilloscope in electronics.
While sequencing is incredibly useful for some applications (I spend my time working on Illumina current-generation "Next-gen" sequencing), we don't understand nearly enough about the regulation of genes for a sequencer to be analogous to an oscilloscope. It is, at best, a wiring diagram, and really much closer to a parts list. Oxford Nanopore seems really exciting, especially for field biologists, but it's at best 1 order of magnitude better (in terms of cost, speed, and sensitivity) than current technology; historical trends have seen a Moore's law-style exponential increase in capabilities, though, so Oxford Nanopore isn't really a quantum leap.
Isn't RNASeq more like looking at really verbose log messages? Where you only sample 1% of the log messages and may be looking for an event that happens 0.01% of the time?
The general assumption is that if you're looking for a transcript that falls below the detection threshold of RNAseq, then it's likely to be so weakly expressed as to be biologically negligible. Furthermore, with the bursty nature of gene expression, an average of 2-3 copies per cell (about my limit of detection in a reasonably sized experiment) could still have a sizable fraction of cells with none at all.
That said, the downside of the Oxford Nanopore for RNA-seq is that, while you get longer reads, it's not yet really clear how many reads you get, which is at least as important for trying to find those moderate-lowly expressed genes.
As a founder of one of the companies mentioned (Ginkgo BioWorks), I have to say this is actually a very nice outline of the industry. It's not clear yet if there's enough of a market for individual services to be broken out and put in the "synbio cloud". Not even clear if there is enough of a market for platforms like Ginkgo that deliver a whole organism to spec vs application-focused companies. Early days, but it's a great field for those interested in getting out of the iphone app rat-race (no offense to the author). We hire lots of programmers.
Does anyone have any good ideas on how we might mitigate the negative consequences of this technology?
I recently dismissed the problems of people printing firearms in a 3d printer as not really being much of a threat as currently firearms are already ubiquitous. However having a potential personal bio-weapon design lab on your desk, would seem to suggest were are entering a whole new paradigm in dangers to humanity and our biosphere.
As an example and a serious question: why won't someone eventually create a virus tailor made to target their enemies? Whether that's an entire race of people, an individual or a family.
The genie is definitely already out of the bottle so I guess the only solution is better defences. We're going to have to get good at identifying threats (possibly in real time) and be able to create effective solutions near instantaneously. Or isolate ourselves in self sufficient habitats as a precaution.
In my mind's eye I can see Bill Joy tapping his watch and saying "Any minute now ..."
> Does anyone have any good ideas on how we might mitigate the negative consequences of this technology? (...) The genie is definitely already out of the bottle so I guess the only solution is better defences.
Agreed. Rob Carlson [1] argues in "Biology is Technology" [2] that regulation can't form an effective means to mitigate threats by limiting access to skills or materials, and that we should instead invest in the ability to develop rapid responses.
Further, he argues that the costs imposed by regulation on skill and material acquisition would weight most heavily on small firms, where most of the industry's innovation originates, thus slowing the very progress we need.
I highly recommend the book, as a review of recent (as of 2010ish) SB developments, comparison to other engineering disciplines, and for a treatment of how the regulatory/IP landscape could best support innovation in biotech.
Before reading those, and before you think to attribute all plagues and disease or suffering to the non-institutional biology crowd, I would like to point out that the world is already full of deadly viruses. The "natural" death rate from (for example) influenza is >0. This is already a huge problem, but not one caused by the "negative consequences" (presumably of developing technology to work with biology) you alluded to.
Edit: also, connor needs to stop spamming synbiota stuff all over the place
I accept that the world is already a biologically menacing place however it isn't yet full of all the virus that are currently being conceived in my head.
For example I might decide I really don't like the Jews (hypothetical). I could probably find out what DNA is common to most Jews, design a virus to target this commonality and couple it with something nasty that already exists, or that I just invented and release it into the population perhaps by infecting my gentile self and booking a flight to Jerusalem.
Your tone suggests that you're a bit bored by this conversation. Are you yourself satisfied that the risk of disaster is acceptably low?
> Your tone suggests that you're a bit bored by this conversation. Are you yourself satisfied that the risk of disaster is acceptably low?
Sorry, my perspective is very unusual. First, whether or not a nation-state bans biology (perhaps because of some precautionary stance) will not dramatically change whether or not I continue to work on technology development. Second, I already find the current level of death and disease unacceptable, and risk calculations don't help me fix things. I'm just completely uninterested in risk calculations.
Edit: I was being somewhat unfair to your original post, which I had judged to be more alarmist than it actually was. I agree that mitigation strategies are important.
I recently dismissed the problems of people printing firearms in a 3d printer as not really being much of a threat as currently firearms are already ubiquitous.
Your reasoning on guns was correct. Remember how almost every year we worry about what kind of influenza we'll get? We all worry about a repeat of the 1918 flu, which is quite likely on any given year.
Remember how antibiotic resistance keeps getting worse and worse (or better and better from the bacteria's perspective)?
HIV originated in nature, probably somewhere not too far from where Ebola is natural brewing now.
Our number one threat is mother nature. Her and other humans living close to domestic and wild animals around the world and air lines.
At any given time we might have to cope with an 1918 like flu.
Custom tailoring a virus to include one people but not other is going to be extremely hard. We are all very closely related. Human conflict tends to be between neighbors, and for example the people most genetically similar to the Israelis are the Palestinians. So the probability of a custom tailored virus is far, far, FAR lower than the probability of a natural don't-give-a-crap-who-it-kills virus.
it seems that Ashkenazi jews are quite distinct from Palestinians. there might be a small overlap between the groups(upper right corner), but a terrorist could still target a specific subgroup of the ashkenazi jews. and using more principal components might get him a better separation between the groups.
I'd disagree, I think it is a lot easier to create a virus that wantonly destroys life than it is to create a cure that puts everything back together. I could email out a script that just executes "sudo rm -rf /" on everyone's computer, and it would probably ruin a lot of people's lives.
Additionally, cures by their nature will always lag behind diseases, people won't invest resources curing a disease that doesn't exist. A non-trivial amount of damage would occur before society reacted to the threat.
The parent's reasoning would probably apply more to a vaccine than a cure.
Also, it's not illogical for there to be a demand for (research into) protections against threats which have not materialized yet, but seem to have a chance of doing so. As the prospect of portable bioweapons becomes more likely, there will be increased incentive to come up with protections from those weapons, even if they have not yet reached viability.
Great question! - I'm glad that folks are thinking about safety, and that it's part of the conversation here.
Earlier this Summer I was invited to attend the FBI/DIYBio outreach event in Walnut Creek, CA. (Disclosure: I'm co-founder of http://synbiota.com - we provide web-based tools and crowd-innovation environment for SynBio) Of course, one of the biggest take-aways from this meeting is that very special attention needs to be payed to how we manage safety.
The simple fact is that right now the ability to create a "Select Agent" (catch-all term for bad biological things) from scratch using this technology is currently out of reach of anyone in the DIYBio scene. That said, it may be just a matter of time until technology gets to the point where it is possible.
There are positive uses for having access to DNA code for a Select Agent, e.g. creating a bio-sensor that will warn us against contamination, or creating an antidote to the offending agent, so it may not be in our interest to completely ban this DNA code.
The good news is that in SynBio we've not created anything from scratch - all that we know comes cribbed from Mother Nature's recipes. Knowing this, each biologic has it's own DNA signature that can be used to screen out any potentially dangerous projects before they are synthesized.
This is just scratching the surface of this question. A great resource to learn about the ethics surrounding SynBio is the SynBio Project by the Woodrow Wilson International Centre for Scholars: http://www.synbioproject.org
Interesting side: When we were at the FBI/DIYBio outreach event we all received a deck of Pokemon-style "Select Agent" playing cards, which are pretty unique. We have some pics of them posted on our blog: http://synbiota.posterous.com/tag/playingcards
Some people think these cards are a bit creepy, while other's can't get enough of them! I'd love to hear HN's thoughts on this!
We actually have created a few things from scratch. Specifically I'm thinking of Top7, a protein that a friend of mine works on: https://en.wikipedia.org/wiki/Top7
The big gene synthesis companies are all a part of the international gene synthesis consortium, which has loose set of guidelines that include gene sequence screening (eg: every order is screened against a database of known pathogens, toxins, viruses, etc.).
granted this is a hard problem (similar to the difficulty of fingerprinting computer viruses). So while synthesis prices are still relatively high, simple fingerprinting will probably suffice for now. There will probably be a lot of development of more advanced machine learning algorithms to detect dangerous sequences as it becomes more and more of a threat in the future.
> The big gene synthesis companies are all a part of the international gene synthesis consortium
Yes, they screen against one of the UK's lists of unauthorized individuals. At least, this was according to Howard Simon (dna20) when he mentioned it at the FBI/DIYbio workshop.
I was curious to see if anyone from DIYbio was on this list, so I did what any self-respecting hacker would do and wrote up some quick code to check... here's the details:
It's good to know that some thought and consideration has gone into safety. Though this does just seem like a temporary measure that will keep the white hats in check. The goals of a lot of these companies is democratization. Would you agree that all the steps of the organism creation process will (in short order) be achievable in private, for good and for ill?
Personally I do think we will need some sort of active technology for protection and we shouldn't put much stock in the efficacy of legislation. We haven't been able to legislate away criminals, and we only need one criminal/virus live in the population to do an inordinate amount of damage.
One defence that is already in place occurs at the DNA printing level. Because of the cost of this equipment, most DNA printing is outsourced to third parties. These guys compare the DNA you send them to that of known pathogens like Anthrax and if your DNA matches they won't print it. There are many other people thinking about this space and what other safeguards we need, one of the main ones is the early detection system you talk about... with cheap DNA sequencing on a future iPhone maybe this will help!
The scary part is not the regulated market, but rather the black market. What happens when an unscrupulous mafia family starts doing DNA printing-for-hire?
Where does an engineer/programmer start to come up to speed on synthetic biology? Are there any essential books in the field like "Learn Bio-hacking in 21 Days", "Bio-chemistry for Dummies", or "The Structure and Interpretation of DNA"? Or do those still need to be written?
It's not on synthetic biology, mind you, but presumably some understanding of the natural biology is a prerequisite to doing the synthetic stuff. Anyway, he works for GSK doing some pretty interesting stuff and is an all around smart guy, so I tend to assume his recommendation is a good one. YMMV.
I too think this is something to be on the curve of. You're gonna have to at least learn the basics of [organic] chemistry, genetics, molecular genetics and cellular biology. A blackbox approach will give you a very low ceiling.
Some books to start with:
Genes, Proteins and Computers
A Practical Introduction to the Simulation of Molecular Systems
Ron Weiss at MIT and Adam Arkin from Berkeley are working on a textbook which will ship sometime this year. But yes, there is still a first textbook still to appear for this field
One of my favorite molecular biology books of all time is "36 Lectures in Biology" by S.E. Luria [1]. Luria was a great biologist and he wrote this book in a colloquial style based on Lecture materials from his MIT course in the 70's. It's accessible but deep and includes hand-drawn line illustrations and several appendixes on thermodynamics & chemistry to get newcomers up to speed.
It provides an unpretentious, solid introduction to core molecular biology theory, written in a personal manner by one of the 20th century's greatest biologists. And it's only $5 w/ shipping on amazon.
Read it, then go watch some of the short videos [3] by iGEM teams on their projects to get a sense of what's made possible by the synthetic biology approach.
The best textbooks for this (this is something of a consensus in the field also) are Molecular Biology of the Cell by Alberts et al. and Lehninger Principles of Biochemistry by Nelson and Cox.
The above doc aims to cover the basics of molecular biology projects in non-institutional environments. I suggest finding your local DIY group and hanging out with them for a while:
I'm a Co-Founder of Synbiota a startup in this area. We've shipped a beta of GENtle, an open-source, web-based (HTML5) DNA editor that you can try right now at:
We're also in closed beta of our integrated environment for collaboration, project management, file storage and DNA synthesis ordering etc. We have a number of iGEM teams as well as some DIYBio groups that are currently using the platform in their everyday work, which is pretty cool ;-)
We've been a part of Mozilla's WebFWD accelerator since last Fall which has been a huge boost for us. Definitely check them out if you are working on a web-standards based startup!
We also recently launched the S Prize - our contest to incentivize the development of open-source plug-ins for GENtle (the DNA editor) - we have a cash purse of $5k on the line for any developers who submit the most useful and delightful contributions.
Ahh, forgot to mention; we've also recently partnered with Genomikon (http://genomikon.ca) to offer biological kits (think Arduino for bio) through Synbiota (all documentation and DNA code is hosted and editable on the platform).
So far we've only sent out one kit to a DIYBio group (still finalizing up production), but we hope it's the first of many.
Transcriptic (https://www.transcriptic.com/about/jobs) is definitely another company to watch in that space, and they are hiring. Hopefully the legions of unemployed bioinformatics and cast-off pharmaceutical researchers will take their expertise to these new enterprises and catalyze this nascent revolution.
The most part of synthetic biology is a brick and mortar science which will need wet lab bench and fiddling fingers to do the development of the prototypes to test the hypothesis. So Synthetic Biology in its entirety is not going to cloud.
Synthetic biology involves several steps or components such as designing a gene,synthesizing it, cloning it into appropriate vector and then expressing it. Then studying the expressed gene product in the presence of multitude of other gene products in an unknown environment poses it's unique challenges.
However several components of Synthetic biology can be moved to cloud. Take an example of designing a gene of interest. It can be done taking a vast amount of data/information of several similar genes from clouds and analyzing and then optimally designing the gene of interest.
The second example is testing the effect of the gene or its product in silico ( on a Lab-on-a-chip) before going to test it in real animal models thus saving costs and time. So in-silico simulated lab animal software models comprising of complex metabolic pathways can be stored in cloud for analysis from several researcher groups or any groups across world. And this is just the surface I am trying to scratch.....
The wetlab component of synthetic biology seems unnecessary when a lot of that work can be automated, which is where this idea of a cloud for biology has merit. Of course if you can do accurate simulations as you describe, that would be the best route.
I like how this summarized the state of the industry, but I feel like "the cloud" angle was a bit forced for a lot of those companies. Most of these companies had nothing to do with the internet / cloud.
Hi, this is Omri, the founder of Genome Compiler (http://genomecompiler.com) mentioned in the article. It's great to see many of my fellow entrepreneurs in this field shamelessly plugging away their companies :-)
Anyhow, this is indeed a great field to be in and I'm happy HN is taking notice.
I have no idea how synthetic biology works, but have a couple of questions from the applications section in the link.
They talk about designing vats of bacteria that extract gold from water. If this level of customization is possible, is it also possible that you could use this to create new disease strains ? For example, can you take a Tuberculosis and engineer it to make it harder to cure? Or create entirely new diseases with elements from different infections?
yup that's one of the biggest fears with respect to bioterrorism.
engineering bacteria to produce new strains or even deadlier toxins is well within realm of synthetic biology (although currently extremely nontrivial).
Falling synthesis prices and better assembly techniques however will make the repeat-test cycle fast enough to debug and engineer such organisms.
Another thought is: how "wet" are a lot of the equipment being sold/marketed to "home"/"bedroom" synth biologists? IIRC from college days, there was a huge overhead in terms of, not only the apparatus (which I gather is coming down in cost), but also the chemicals in use, which require all sorts of safety regs, OSHA compliance, etc etc. How much of a barrier is this nowadays?
i remember ginkgo bioworks in the news, along w/ the MIT mail order biobank. Any pointers onto papers that have come out of this or new medicines, etc?
Has any synth organisms other than what's come out of Venter's lab been created and proven viable?
>Has any synth organisms other than what's come out of
>Venter's lab been created and proven viable?
You should think of the Venter work as a technical demonstration of building a very large piece of DNA (>500 kilobases, a small genome) and transplanting a whole genome into a microbe that had its original genome removed. Most of the immediate practical applications of synthetic biology don't need pieces nearly that large. If you want to add a genetic pathway to produce a chemical product, it's typically <10 additional genes that need to be added to a production microbe and that will be <50 kilobases typically.
> Any pointers onto papers that have come out of this or new
> medicines, etc?
There are lots of industrial microbes out there now. For example, most Amino Acids that go into food supplements are made at massive scale - >1M tons/year by engineered microbes.
Seriously cool and scary technology. "many of the startups are going for "killer apps" first." he wrote this with no irony intended! I'm considering getting a degree in "bio-informatics"
It's pretty cool to see Synthetic Biology getting coverage on HN. One of the really interesting ideas that scientists are starting to grab a hold of is that of "biological parts". People have started compiling these biological circuit elements and figuring out how they work when exposed to different stimuli [1]. The ultimate aim is to standardize the usage of these parts so that you could design an organism in your browser, figure out whether it will work or not (based on a super-secret algorithm), and get the resulting functional organism shipped to your lab bench in a couple of days. As a reference, right now it can take up to 2 months to even create an initial DNA product with traditional methods. From there, it could take months to get your organism working correctly. New DNA synthesis techniques will dramatically speed up some of the initial steps (something Teselagen has done).
An addition to the article I'd make is that Teselagen is actually letting people leverage new DNA synthesis techniques produced by a couple of groups (the Venter institute is one) to design better biological circuits (as in faster, better, stronger). If you are interested in reading about how scientists have begun to overcome some of the many barriers to biological circuit design, there's a good intro to advances in DNA synthesis techniques here: http://j5.jbei.org/j5manual/pages/1.html
[1] partsregistry.org