Fog Creek Software
Discussion Board




Leveraging Computer Sci Skills in a Biotech World

Everywhere I look, I find articles claiming that biotech is the new hotbed of innovation, and will be the place to be  during the jobs rebound.  Trouble is, I majored in Mathematics and Computer Science - with not a single Biology course since freshman year in high school.

Will you attempt to leverage your computer science skills and cross over into the biotech world?  If so, how?  Will you get further education in biology or biomedical engineering or something? 

Adam N
Tuesday, February 10, 2004

"Will you attempt to leverage your computer science skills and cross over into the biotech world?"

I would but Im still using a tactical approach to creating synergies through leveraging the vertical efficiencies created by owning a website.

maybe Ill find some time to cross over next year :)

FullNameRequired
Tuesday, February 10, 2004

I majored in molecular biology at M.I.T.  When I lost my last job at the end of the bubble (circa July 2001) I applied for a heck of a lot of computer oriented jobs at a heck of a lot of biotech companies.

No responses.  None.

Things may be changing especially as the job market comes back but I don't think I would pin my hopes on "I just need some bio training and I can make a killing in this new genetic engineering stuff".  It was new in the 1980s.  Now it's just useful stuff.

name withheld out of cowardice
Tuesday, February 10, 2004

Become a consultant installing LIMS (Laboratory Information Management System).  There are quite a few variants out there, but everyone needs one and there's a real lack of people with solid database experience and logic training who are able to set one up.

And it's not too boring either, and you aren't limited to BioTech; you can travel to all the great pharmaceutical hotbeds like Philadelphia and San Juan.

Lou
Tuesday, February 10, 2004

Perhaps of interest:

http://www.business2.com/b2/web/articles/0,17863,571720,00.html?cnn=yes

"As you'll see, many of today's most exciting startups are rooted in the well-funded fields of medicine and biotechnology..."

bob
Tuesday, February 10, 2004

what's bioinformatics and does it have anything to do with biotech?
888

anon
Tuesday, February 10, 2004

Bioinformatics is the use of computers to handle biological information.  Think big databases and huge text documents  full of the letters T,C,G & A.

a cynic writes...
Tuesday, February 10, 2004

And processes surrounding the generation of such data.  Want to find out why Joe Smith's samples always come back tainted - look at the culture samples you've been taking after gowning procedures and see if there's a trend.

Part of it is data, part of it is meta-data.  It's relatively interesting, and there's a need.  If anything it's a way to start the transition.

Lou
Tuesday, February 10, 2004

http://bioinformatics.oupjournals.org/

Fernanda Stickpot
Tuesday, February 10, 2004

My feeling on this is that computers are computers and code is code. Whether your field of endeavour is biotech or not, a CPU is still a CPU and RAM is still RAM, and they fit into the same mainboard that hosts the same video and network cards. When you write code, it's still the same if's and for's and declarations. Programming languages are the same way; as long as I can write a conditional and a loop, I can probably hack my way through whatever needs to be done, no matter what language happens to be used.

The critical angle, IMO, is whether you can talk productively to biotech people. The real job of most computer workers is to translate requirements into a language the computer understands. When a biotech worker says he needs something that can display the full chromatographic profile of some specific chemical, that should be all he needs to know, and you should be able to figure out what that means and fix it. All he knows is he can't do what he needs. If it's because his CPU is too slow, his display is too small, or his software package is too limited -- that's not his problem. That's what *we* are supposed to know and fix.

Caliban Tiresias Darklock
Tuesday, February 10, 2004

Caliban:

I disagree.  There are tremendous advantages to programmers who also know a great deal about the domain they are working in.  The communication between the domain expert/analyst and the (possibly quite expert) coder is on of the larger bottlenecks in programming.

If I were hiring for a financial application, for example, I would be really impressed by a good experienced coder who also had an MBA in finance.  With understanding of the software and the domain he should be able to come up with novel software solutions unimagined by domain or software experts alone.

name withheld out of cowardice
Tuesday, February 10, 2004

I talked my way into a bioinformatics job with a math and CS background about 2 years ago, and recently left.  "Biotech" is very broad, and there is innovation going on - just as there is inovation going on in the field of "computers". 

There isn't a job boom in biotech like there was in IT a few years back. In fact if there ever was a biotech boom, it has now busted. It is pretty hard to get a job doing something interesting even if you have impressive credentials.

As someone mentioned, bioinformatics is usually defined as the use of data processing algorithms to dig around with huge sequence databases of genetic data. It can also mean visualizing protein folding, or creating graphical maps of biochemical pathways, or using matlab to figure out ways to use signal processing to reduce noise in microarray readings.

Bioinformatics is the easiest field within biotech to get into if you have a strong programming AND quantitative background. Stress on quantitative. If you just have a strong programming background, you will be stuck doing maintenance programming for some sloppy attempt at porting a PERL script to JAVA or .NET or whatever - there are probably better jobs outside of bioinformatics for you. If you have a good math background, particularly in statistics and signal processing, or search algorithms, the doors open to do more interesting projects. The more you know, and if you can apply what you know creatively, the more potential there is for you to do interesting work.

Bioinformatics is just like finance, or even computer science in this regard. If you just want to manage the database at a bank, you just need to know how to program and how to deal with database administration. But if you want to work on the quant desk figuring out wizardlike methods for doing extreme derivative arbitrage, you need to not only be able to be a wizard in C++, but a wizard in MATLAB, and have a PhD in applied math or physics. 

In Bioinformatics, if you want to work on mundane stuff, like build scripts and maintaing the local BLAST installation, then you could get into a uni lab or bioinformatics firm with just a CS degree. If you want to help doing more modelling type stuff, a very strong math background, and ability to drive MATLAB is very helpful. A molecular bio course helps. If you want to start theorizing yourself, you need your computer and math skills plus an advanced degree in biology. If you REALLY want to do cool stuff, you need an advanced degree in structural chemistry, or chemical physics, be a wizard in C++, picked up all the molecular bio credits you could along the way, etc...

The problem with bioinformatics is that there isn't a huge market for the software. If you write a palm pilot grocery list application, everyone who ever goes shopping might want to buy it. If you come up with a novel algorithm for locating retrotransposons in the mouse genome, aside from university researchers who expect to get the code free, the only groups who might want your software are giant pharmaceutical companies.

So, your fate is essentially working at a university research lab (not bad work; not necessarily bad pay if you work for a place like Harvard medical school) or working as support staff at a big pharma, or working at a startup company that consults for a big pharma or produces a product to sell to a big pharma. 

That is, unless you go academic, your livelihood in bioinformatics is tied to how well big pharmeceutical companies are performing. 

was once a bioinformaticist
Tuesday, February 10, 2004

> The communication between the domain expert/analyst
> and the (possibly quite expert) coder is on of the larger
> bottlenecks in programming.

I'm going to say this again, because apparently you missed it:

"The critical angle, IMO, is whether you can talk productively to biotech people."

That requires a certain amount of skill in the field, at least in terms of the vocabulary -- but it *doesn't* change the computer skills needed. Whether I'm calculating mortgage rates or organic decay, I will write essentially similar loops and equations and conditionals. But I need to understand the problem, and it shouldn't be the other guy's job to explain it in terms *I* can understand. If I'm working as a problem solver in any field at all, understanding the problem is inherently part of my job description.

Caliban Tiresias Darklock
Tuesday, February 10, 2004

Hi Adam,

I've got a BS and PhD in bio, so I'm coming from the other side, but I've been working in bioinformatics for 3 years now.  I'd be happy to chat if you want to email me.

Reponses to previous posts:
1) Yes, biotech has been in the dumps (economically) for the past few years, lots of companies closing and laying off people (including my own) but a few recent IPOs may mean that capital is flowing back in.  Big pharma is a whole other world from biotech... much less cyclical, as they get $$ from product sales not venture capital

2) There's a difference between computational biology -- using difficult math to solve biological problems, like protein folding -- vs. bioinformatics, which is more like data handling, storage, processing, display, and SOME analysis.  The two overlap, of course, and I do both at my job, but I'd say bioinformatics is closer to IT while the other is more scientific.

3) Take an intro molcular biology course (not general bio) at your local community college.  You won't even get in the door if you can't show you know the difference between DNA and proteins.  On the other hand, depending on the job and seniority level, you may not have to show much more than that.  We have half CS grads here (other half bio grads like me).

4) While you're at it, check if your community college has bioinformatics courses (usually in Perl, alas).  They do here in SanFran.  For $500 or so, you can get an intro the field and see if you like it.

5) It's true there is little market for bioinformatics software or data -- look at Incyte, Celera, HGS, etc.  But it's also true that there are jobs outside big pharma -- in biotech companies.  The usual pros and cons about large established companies vs. startups apply.

Biotech coder
Tuesday, February 10, 2004

*  Recent Topics

*  Fog Creek Home