Slashdot Mirror
Online Artificial Gene Design
massivefoot writes to tell us New Scientist is reporting that researchers at Johns Hopkins University School of Medicine have released a new software suite, GeneDesign, that helps to simplify the steps in designing artificial DNA. From the article: "These key steps include translating proteins and amino acids - the building blocks which make proteins - backwards into a DNA sequence. Or the software can manipulate simulated DNA "codons" which can code for an amino acid. DNA codons are made of sets of three nucleotides - the fundamental molecules which link together to form a DNA chain."
The article discusses how much of the software available today lacks safety checks on the DNA sequences that are produced.
That's really not much different than what we have with many programming langauges today. While most widely used C and C++ compilers today do offer numerous helpful compilation warnings, little is done to verify the safety of the emitted code. Many of the security problems we're dealing with today are due to buffer overflows, and other matters such as that.
It is often quite expensive to ensure software safety, be it when dealing with programming language compilers or DNA "compilers". Indeed, more study will be necessary to determine what a feasible trade off between the two is.
Cyric Zndovzny at your service.
Sure, most college grads trained in molecular biology can cook up their own bio-cocktail of Epogen, Herceptin, Enbrel, etc.
But the big question is
1) would you use your own untested (or minimally tested) cocktail if your life depends on it? ie if you have a cancer that can be treated by the real drug, would you trust your basement concoction to actually save your life?
2) All of the biologic drugs are injected in to skin/muscle or infused by vein. Would you trust injecting your cocktail into your own body? Do you know that the cell lines you used are free of bacteria, contaminants, retroviruses? ie do you trust that your concoction won't kill you?
I suffer, i would say, from duchenne, the first decease dicovered by dna sequencing.
This causes muscle loss, and starts at early adolecence and ends in death between 20 an 30 years of age.
I'am currently 24 and hoping for a cure.
Keep going
What exactly is the exciting news here? This type of software has been around for many, many years. Analyzing a gene sequence to determine restriction enzyme sites, or optimizing codon usage for efficient heterologous expression is absolutely routine, and is performed even in undergraduate level molecular biology courses. It's laughable that the ability of this software to "...manipulate simulated DNA 'codons' which can code for an amino acid" is being touted as an advance.
I can't even believe that New Scientist is reporting this, let alone Slashdot. There must be at least 100 other tools which perform the same functions, many of which are free (both as in beer and source code).
The underlying science is pretty trivial, yeah. (Or at least "well-understood.") But having this tool in one place, as a reasonably well-designed Web app, is neat.
... I think the real thing that bothers me is, why is the biology field so devoid of computer people?!
On to the bigger question
Stereotyping here -- it's a bit of a culture clash. Until fairly recently, biology (with exceptions for some subfields such as ecology) was, to put it bluntly, the science you went into if you wanted to do science but weren't very good at math. And I think it's fair to say that most "wet-lab" biologists still think more qualitatively than quantitatively. They're very, very good at describing things; they're not so hot at putting those descriptions into numeric or algorithmic terms. And, still stereotyping, CS people tend to be exactly the opposite: "if you can't code it, it doesn't exist," and they're uncomfortable with the inherent, um, gooiness of living systems.
Computers are always supposed to behave predictably. Living things never do. It's really that simple.
You also have the opposite problem, overenthusiasm, which is born out of the same kind of ignorance: biologists who think that they can throw a bunch of random microarray or PCR data at someone's analysis algorithm and get The Answer, and computer scientists and mathematicians who take Bio 101 and think they know enough biology to interpret the answers they get. In both cases, of course, both sides are severly underestimating the complexity of The Other Guy's chunk of the problem.
Don't get me wrong; I do think it's getting better. But even someone like me, who's had one foot in each camp for a number of years now, has to admit that we've got a long way to go before quantitative biology really exists as a unified field.
The correlation between ignorance of statistics and using "correlation is not causation" as an argument is close to 1.
its not so trivial to design a gene, actually; if you look around, there is a lot of interaction between CS and bio folks, especially at Hopkins. (see Applied Math/Biophys/BME/EE/ChemE/Bio departments)
There was a symposium last week at the Med School featuring, guess who, Jef Boeke and Drew Endy, among others, talking about High Throughput Biology. Because of the overwhelming numbers of genes and gene products, you really need to start to use some automation and machine learning algorithms to help you predict what's going on with gene regulation and cell signaling networks.
I'm actually ashamed that you, a Neuroscience major at Hopkins, posted something like this. Please, go back to studying cell bio.
There have been innumerable bioinformatics sites put up over the years by various institutions, but they've never been examples of refined usability. You could say that is because the focus has been on the underlying tools and not the interface.
Despite being highly educated and working alongside a great deal of computational effort, biologists in general are not anymore computer literate that your average office worker. Much of the computational side has been taken up by computer literate Mathematicians, Physicists and Comp. Scientists. Those that can bring together a multitude of tools (often rudimentary academia quality software) to solve a problem through programming, find many of their collegues cannot or will not do the same. The older generation directing the labs then ask those that can to design sites to provide tools for the illiterate but as is the case in research, these static tools generally do not provide the necessary power or flexibility. Many efforts end up being ignored in favour of simply casing down the programmer in question.
Ultimately, the best solution is to teach programming to biologists as a core subject. Our old categorization of the sciences is out-moded.
You're new to slashdot, huh? As a biologist/bioinformatics guy, every time I read articles on slashdot that involve my field of research, I see that 90%+ of the 3+ or better comments are crap.
This leads me to believe that in areas that are not my speciality, slashdotters are equally full of shit. Sure, it's just a hypothesis right now, but I'm sure with a little help I could gather convincing evidence...