Bioinformatics in The Economist

Erich Schwarz writes "Bioinformatics has gone from being an esoteric sub-field to being a business. The Economist gives a useful overview, while warning 'Bioinformatics is not for the faint of heart...'"

Comments

What I'd major in

[SteweyGriffin]

If I had to do it again, I'd definitely choose biology or bioengineering or something related.

It seems as most everything in computer has "been done", and biology/chemisty/biochemical engineering seems to be where all the fun & excitement is these days.

Anyone else agree? Just curious.

Re:What I'd major in

[Ed Avis]

Digital computers have (mostly) been around for less than a hundred hears. Biological organisms for several million times as long. In which field is it more likely that 'everything has been done'?

Re:What I'd major in

[Anonymous Coward]

It seems as most everything in computer has "been done", and biology/chemisty/biochemical engineering seems to be where all the fun & excitement is these days.

What about quantum computing? Chemical computing? Optical computing?

Re: What I'd major in

[Anonymous Coward]

I majored in Computer Science because I liked the thoughts and perspective involved in programming, not because I was looking for the next hottest thing. I would like to think that people study things that they are passionate about, rather than things that happen to be new and full of potential.

Re:What I'd major in

[coloth]

I think the amazing thing is how these two fields are merging.

Whether you look at artificial life, prosthetics, advanced protein sequencers, electronic "noses", even robotic pets, it seems like at some point what we think of as electronics and computers, and what we think of as genetics or biology are all going to be part of a continuum.

I don't think it matters what you majored in--these days you can cross over.

Re:What I'd major in

[El Pollo Loco]

No, I don't agree. I don't agree at all. We are on the cusp of the biggest improvents in computing in my opinion. It's like cars. First 50 years there was huge improvements. But they didn't become really useful, commodity goods even, until everyone had one. Then the real useful work begins. Refinement. Same with computing. The largest gains in terms of speed, etc have probobly happened already. But now we have the golden age to look forward. Where computers, as useful as they are now, will become more useful(kinda redundant, but oh well). We will soon be able to have computers that don't crash, that work with everything. Moreso, everyone will be able to use them, like cars. And then, when most can afford to use, and most can use, is when the real fun and excieting stuff happens. Computers that do the work, instead of you doing most of it. Photorealic games, surgey from half way around the world. The golden age is coming.

Re:What I'd major in

[F2F]

totally agree with you. computer science is not the forefront of technology anymore -- it's too commercialized. or maybe it's just too complex these days (always a good read -- "System Software Research is Irrelevant", a talk by rob pike that summarizes most people's feelings towards operating systems and computer science in general).

the next big thing (speaking from the point of view of someone who has spent the last 7 years in a north american university) is biotechnology...

of course i hope i'm disproved -- i like computers, not counting genes...

Re:What I'd major in

[coloth]

The golden age is coming.

I agree, and I'd just go one step further. I don't think computing will ever reach a point of stagnation. Computing is about ideas, and as long as we are human, we'll always have new ideas.

Re: What I'd major in

[Anonymous Coward]

Fuck you. What if I happen to be passionate about medieval sailing ropes? What kind of job will I get? Will you pay for me to live somewhere because what I'm passionate about doesn't pay?
This society that we created forces people into things they don't care about so they can eat.
You happen to like computers and it pays right now. So get off your high horse and consider the point of view of other people.

Re:What I'd major in

[Bicoid]

It seems as most everything in computer has "been done", and biology/chemisty/biochemical engineering seems to be where all the fun & excitement is these days.

The difference in modern computing and modern biochemistry/genetics/developmental biology is that bioenginnering is a good deal behind computing. 50 years ago, we were working on making computers work. We didn't really get to working on all the applications of computers until computers could be personalized and used in the home. Bioengineering is still at the point where it's being explored. It's all being done in labs because there aren't "do it yourself" genetics labs you can put in your home. Eventually there will be, but not right now. Therefore, if you want to fark around with DNA, you need a degree and a job. If you want to fark around with computers, all you need is to teach yourself some languages and get a PC.

It's not that computers aren't as exciting as genetics is right now...there's a lot more you can do with computers than DNA at the moment. However, genetics is more exclusive because it is not as availible to the public.

Re:What I'd major in

[BWJones]

I don't think it matters what you majored in--these days you can cross over.

When it comes to bioinformatics, this is certainly true. However, to be successful in bioinformatics, people need a basic knowledge of chemistry, biology, molecular biology, genetics, programming, medicine, comp. sci. and perhaps some library science to be proficient in bioinformatics and this combination can be hard to come by. Additionally, an ability to successfully communicate with folks from all of those disciplines is crucial and can be even harder to come by. For another article on some of these issues in bioinformatics (with an admitted bias towards the Macintosh) read this article I wrote for Scientia et Macintosh

Re: What I'd major in

[s20451]

I majored in Computer Science because I liked the thoughts and perspective involved in programming, not because I was looking for the next hottest thing.

Wouldn't it be great if you could do both? Wait, you can: a lot of the great bioinformaticians are computer scientists (through fields like machine learning).

Re: What I'd major in

[czarneki]

Some people are passionate and excited about a field precisely because it is new and full of potential. I think there's nothing wrong with picking your concentration based on its "newness" and the potential to make important advances and discoveries yourself.

E.O. Wilson wrote in his autobiography that he picked ants to study because no one else was studying them -- and he freely concedes that he's probably not as brilliant as some of the other biologists, but he picked a field that was new and exciting and he ended up having a lot more satisfaction as a result because he could make important contributions personally in a field that was still sparsely populated.

Re:What I'd major in

[coloth]

this combination can be hard to come by

I guess there'll be a few with outright credentials, some more who are smart enough to wing it while they learn what they're missing, and a lot of people who contribute as part of a team.

If you haven't read Ship of Gold by Gary Kinder, about Tommy Thompson's life and eventual recovery of $1 billion in gold from the deep sea, I strongly recommend it, especially to any scientific or engineering-oriented person. Tommy's story is a case study in how to break new ground, think creatively, and form networks of expertise. It convinced me that credendials, and even past experience, may not be the best indicators of future success.

Re:What I'd major in

[DAldredge]

And the current big idea in computing is "Move it all to India and damn what it does to our home country!"

Re:What I'd major in

[GuySmiley]

As a molecular biologist, I steer interns out of biology. Blend bio with any other discipline. You will spend your entire life justifying the 90% of experiments that fail, justifying your existance to get grants, etc. Bio-business, bio,-programming, bio-sales, bio-anything. Lab rats at the bench make no money, get zero respect, get zero creative control and are this generation's ditch diggers.

A decade of scientific failures is indescribably difficult to explain, but it is ungratifying. I have been taking Linux/SQL classes for the last few years trying to steer into the digital world, and thus mine and slashdot's interests are now intersecting.

Yes, having interesting work can only go so far. You still have Senior management flushing the cool stuff you invent down the toilet or left on the vine to rot. At the end of the day , you still have a pile of bills to pay.

looking for meaningful work in MPLS fleabag-at-covat-dot-net

Re:What I'd major in

[coloth]

and your home country would be... Pakistan?

Re:What I'd major in

[t0rnt0pieces]

As a molecular biologist, I steer interns out of biology. Blend bio with any other discipline. You will spend your entire life justifying the 90% of experiments that fail, justifying your existance to get grants, etc. Bio-business, bio,-programming, bio-sales, bio-anything. Lab rats at the bench make no money, get zero respect, get zero creative control and are this generation's ditch diggers.

I couldn't have said it better myself. Here's something to chew on. Do you need a bio degree to do bio-sales, bio-business, bio-programming, etc? Most likely not. Biologists generally know nothing about business, sales, or programming, so they can't get jobs in these fields.

I work as a lab tech and I agree that they are ditch diggers. I actually did go to grad school for a bit, I pretty much hated it. It's very unsatisfying. At least if you program you can design something *useful* that *works* and feel like you've accomplished something. Biologists go months without accomplishing anything, and the work is dirty. Who wants to be in a lab for the rest of their lives? Even the PhD's have it really tough. No one hires PhD's, they're overqualified. Most of them end up working long hours for years as post-docs making $30,000.

Re:What I'd major in

[GuySmiley]

Right on

I just hired a kid with a Masters degree to wash dishes. It may be today's work environment, but I see this all the time, especially at the PhD level, people wanting /any/ non-academic job because they are tired of being pissen upon. (pardon my run-on)

I love my work, but I hate my job.

Comprende?

Re:What I'd major in

[Goalie_Ca]

I'd have to agree, even though I"m only in first year. I intend on choosing the biomedical stream along with computers or physics option. That stuff is really kick ass!

Re:What I'd major in

[furchin]

As a biology and computer science major (and a master's in computer science) from a top 25 school, I feel I have a right to speak my mind about bioinformatics. Bioinformatics is getting a lot of attention from the media, and you'd be forgiven for thinking that it is new and exciting. Really, the reason it is getting so much attention is because we just recently completed drafts of the human genome sequence. Thus we have a whole bunch of data, and don't really know what it means. Bioinformatics simply takes things we know (ie, the sequence AGCCTG_GT is a promoter sequence) and searches a huge long string for all instances of that sequence. Or we're trying to locate repetitions of one region of the sequence in another region. It's really not that exciting. Yes, it is really important to us since without these DNA strings, we wouldn't be here, but its just string matching on a grand scale, and nothing more.

Re:What I'd major in

[t0rnt0pieces]

People go into biology because it is the easiest of sciences( slightly easier than Comp. Sci, if you can call that a science). I thought it was understood apriori that this is what biologists end up doing, and doing it with a very low, low salary.

It depends on the individual as to what they consider "easy." I never thought bio was very easy, but I suppose compared to chem and physics it is. Physics majors aren't exactly setting the world on fire with their salaries either. I suppose chem is better, but I don't know much about working in that field. To me, computer science is easier than bio because it's more fun. But I know a girl right now in one of my CS classes who's a bio major (and claims to have good grades) and is failing the CS class. Some people understand it, some just don't.

Re:What I'd major in

[the gnat]

It convinced me that credendials, and even past experience, may not be the best indicators of future success.

Um, sorry, but it's extremely difficult to go anywhere in the sciences without a PhD. I seriously doubt that will change any time soon. The people doing the hiring are biologists, and would not even bother to read the resume of a whiz-kid programmer unless he had some letters after his name.

Re: What I'd major in

[the gnat]

Yeah, I actually work in bioinformatics right now, and I'm going to switch fields entirely to something where I can concentrate on the science rather than talk about what hot shit bioinformatics is. Too many idiots (with and without PhDs) are going to be flocking to the field because they read in "Fast Company" that the field was booming.

Unless you have a PhD in CS or Bio right now or will be getting one soon, I suspect it will be too late to start in bioinformatics. By the time you are considered qualified enough to get a real job, the market will be saturated.

Re:What I'd major in

[Zork the Almighty]

I'm a graduate student in mathematics and I find biology "hard". It's somewhat straightforwards, but there's too much crap to memorize. But all the biology people I know think math is hard, so it's obviously a matter of personal attributes and taste as to what is "hard" or "easy".

Re:What I'd major in

[GuySmiley]

I switched from ChemEng and IT to BioChem for a girl too. Dang. Mod me down for that.

Life is finding a balance between internal gratification and external resource necessity. What you find intuitive now , may not be where your interests (or the market) lie in 15 years. Biology skills also may not be what will be required to support a family in that time also. Heck, I made more per hour parking cars 15 years ago.

Try to develop, in parallel, 2 skilled talents and play them off each other. You will be able to say, " I am a programmer who knows how to run an ABI 3100" or a " a geneticist who can grep any text file with the best of them".

But then again, nobody believes you unless you have your PhD, so it comes down to selling yourself in the end.

Cheers

Re:What I'd major in

[doug renfrew]

As a molecular genetics major, computer science minor, and chemistry minor at the University of Vermont I can say Yes.

Re:What I'd major in

[coloth]

would not even bother to read the resume

I know what you're saying, and I basically agree in most cases. However:

Tommy Thompson, through rigorous thinking, relentless networking, holistic design, etc... Achieved a scientific and engineering feat which far outdid the full resources of the US Navy, Woods Hole Oceanographic Institute, and, of course, countless treasure seeking dilettantes, while spending only a few million dollars from individual investors.

He did not have a PhD, but spearheaded and managed a revolutionary project that led him to employ his former superiors at Battelle, a major defense research lab.

When his team finally found the site of the SS Central America and saw her riches via his revolutionary deep water robot (late 80's), he wouldn't touch a coin before completely photographing and videotaping the site to preserve its archaeological integrity. Each coin of the thousands recovered was individually photographed in place and once retrieved by the robot, while some on the team groused at the delay.

It is difficult to convey the holistic nature of his approach in just a few paragraphs. But I believe his example proves my point that in a new field (in his case, deep water exploration and recovery), precise credentials don't matter.

If somebody thinks you can help him win the Nobel Prize, he'll hire you. If you use letters after your name and a resume to convince him of that, fine. If a 30-minute conversation does the trick--hey, that'll work too.

Re:What I'd major in

[coloth]

YES! Ship of Gold is an EXCELLENT book.

My copy of America's Lost Treasure is on its way from Amazon. For those who haven't read Ship of Gold, what you will crave more than anything as you read it is pictures.

Apparently, Tommy's coffee table book is a near-perfect complement to Gary Kinder's excellent narrative. I must say, Mr. Kinder is a brilliant writer, however some of the scenes he describes strain even the vivid imagination.

As a person with no imagination at all, I can't wait for the picture book.

Re:What I'd major in

[the gnat]

But I believe his example proves my point that in a new field (in his case, deep water exploration and recovery), precise credentials don't matter.

If somebody thinks you can help him win the Nobel Prize, he'll hire you. If you use letters after your name and a resume to convince him of that, fine. If a 30-minute conversation does the trick--hey, that'll work too.

This all sounds great. There are two problems:

1. Bioinformatics is a hot field, but not really a new enough of one to be a true meritocracy the way you describe. It's very important, but there are already many people actively doing research or getting their degrees. You can't just break into the field (unless you already have a scientific reputation). Barring a degree, a lot of publications helps- but that's going to take a while and involve considerable effort.

2. You have to get to the point where you get the 30-minute conversation first. This is harder than it sounds. I was offered only one interview, and landed a job that was aimed at PhDs or really advanced people (I was not even through with college). However, all of the people interviewing me knew my boss personally. So I was aided by my connections, and it took a while to get those. Had circumstances been different, they probably wouldn't have bothered to contact me.

(I ended up not taking the job- it would have forced me to wait much too long to attend grad school, and didn't quite pay enough to make that worthwhile. Though I confess I still wonder if I made a stupid mistake. . . most jobs for people at my level are along the lines of "database jockey" or "web monkey". These people wanted me as a scientist, probably the highest compliment they could have paid me.)

Re:What I'd major in

[Fear the Clam]

Computing is about ideas, and as long as we are human, we'll always have new ideas.

You haven't seen a movies or television lately, have you?

Re:What I'd major in

[irfco]

It's interesting. Why does the combination of 'art' / 'comp.science' work so well? I know a lot of people who have succeeded quite well with such majors. They seem to be a 'natural' combination. Is it simply the way industry has turned in the past few years? Any thoughts?

Re:What I'd major in

[coloth]

You haven't seen a movies or television lately, have you?

That's a good counterpoint, but not a great one.

With movies and television, there are plenty of ideas, but the very limited (and expensive) channels of distribution prevent most of them--especially the experimental ones--from being tried.

In computing, the channels of distribution are much broader and cheaper. For less than a hundred dollars and a few days work, anybody can launch a website. Within a few weeks, it will be in most search engines, and if their information is useful, well constructed, or entertaining, that can leverage them into a whole business.

Or you can develop shareware and put it onto download.com.

Or you can invent a new compression algorithm if you're smart enough.

Or you can work on getting computing resources to the poorest people in the world.

I don't know, it seems to me that computing has a near endless future.

I mean, just think about Linux. You have a worldwide, voluntary community challenging one of the greatest companies in history. That doesn't inspire you?

Re:What I'd major in

[timeOday]

Interestingly, Bill Joy said exactly the same thing on the Charlie Rose show a year or two ago.

Re:What I'd major in

[frenchs]

While you are correct that bioinformatics involves a lot of string matching, you are trivalizing the scope, and the fruits of bioinformatics.

For example, doing an MSA on a set of strings is easy enough, and the computation involved in the algorithim to find the best alignment is slightly interesting. But the point of an MSA is not to parse the file, but to gain new knowledge from a bunch of data that was previously meaningless. And when we have thousands upon thousands of genetic sequences from various organisms, comparing them on a genetic level is definitely "new and exciting".

A Simple Example
For those that know that Humans are 98.7% similar in DNA to a chimp, the question is really why are we so physically different. Well, bioinformatics, combined with lab research has provided insight into this.

We now know what makes us fundamentally different from the Chimps is that the SNPs that we have that make us 1.3% different are in key locations such as a genes which plays a role in developmentmental processes (and particullaraly the brain). For additional information on this, see: "Intra- and Interspecific variation in Primate Gene Expression Patterns; Science 4/13/02"

And in the 8/16/02 issue of Science, there was a short snippet on the discovery of a particular gene, FOXP2, that plays a role in speech and jaw development. They discuss that mutations in this gene have varied, and wide effects on the phenotypic expression in the organism. Now, lets apply what we know bioinformatics can apply, by asking the question. "What difference in the FOXP2 gene causes the differences between humans and chimpanzees in speech and jaw development".

There are SNP databases out there that have been constructed by analyzing the various known genetic sequences and if you search that database, you will notice there *is* a SNP in humans that is unique to humans.... now we have a really interesting argument to why humans can talk and other species can't. (http://www.geocities.com/asdut2002/FOXP2.html

Conclusion
So the bottom line is that yes, bioinformatics allows us to do a lot of string comparisions, but the benefit of those comarisions are the real gem. And without computer scientists to construct algorithims, and IT professionals to develop world class database systems, this information just isn't useful.

Re:What I'd major in

[mlush]

For any computer type looking to get into bioinformatics. I will heavily reccomend you read Larry Gonick's Cartoon Guide to Genetics as it goes from absolute basics to 1st year biology degree

Re:What I'd major in

[metlin]

Well, that's true for most pure science research disciplines. Being a CS researcher, I have to justify my work as being relevant to the industry to get grants etc. Creative control? That's something that the people at the top decide for you.

I guess this is true for a lot of people in Math, Physics and other areas too.

Contrary to what people think, there is little money in pure theoretical CS (in my case, AI) research. The only way to get recognized is to show commercial relevance.

The funny thing is that, publications carry you only so far, unlike some areas like Physics where good publications will keep you alive. I guess the same would be true for Biology too, since like CS it is also more of an application level science. Unfortunate, but inevitable.

Re:What I'd major in

[mlush]

Biology is hard because Mother Nature is the most dirty hacker on the planet!

Imagine trying to unpick the code of a programmer who:-

does not know how to program, so just randomly diddles with the code (any code, the program, the compiler or the operating system she don't care) and uses whatevers just good enough.

who does not know about documenting, or version control

who's software testing is done by the users

and when she does reuse her code, its sometimes the result of two projects converging on the same solution.

Re:What I'd major in

[littleRedFriend]

I work in a bioinformatics company (Boston, MA & Paris, France) and what we do is diverse. From text processing, statistics, graphs, string comparison, database management, data integration, data visualization and more.

We hire people with various skills. All people, including the biologist (or bioanalist, people that generate and analyze biological data with the tools we make), require a thorough understanding of *nix.

Computer science skills that we are particulary interested in are string matching, graph theory, statistical algorithms, super computing (software that run with 1000s of threads). This is all in C and C++.

We work with (real) computers from every vendor (IBM, Sun, SGI, Compaq, ...) and also have Linux clusters. We have good sys admins and custom web application developers. RDBMS and OODBMS knowledge is a pro. Web scripting, like perl and PHP, is a must.

All, people that work here are either computer scientist with an interest in biology or the other way around. Bioinformatics is a lot of fun, but the market is difficult right now (still troubled by the internet bubble investment hype - some big, well funded companies went down).

We get about 20 resumes per week, most of them applying for a bioanalist position. We have enough of those. We do not get many serious CS applications (finding people with some experience or specific skills is difficult), but when we do we seriously look at them.

So go ahead and check us out (shameless plug).

Re:What I'd major in

[baz00f]

Without having a PhD, Jimmy Eng ( Eng bio 1, Eng bio 2) has done a helluva lot for the field of proteomics devoloping the groundbreaking software SEQUEST and related utilities.

Re:What I'd major in

[the gnat]

Good example, but read his resume closely- it helps prove my point. He had access to the right people and was able to work for a long time at a stable position. His connections enabled him to get jobs where people let him do real work. I'm not questioning his talent- his work looks awesome- but he didn't exactly break into the field. And academic environments are very different from industry.

Re:What I'd major in

[proteinaceous]

> Biologists go months without accomplishing anything, and the work is dirty.

True. I can't think of any other field where 90-95% of your work doesn't work or doesn't amount to anything (except as a "learning experience"). The hardest part to deal with is that in this field hard work does not necessarily equal success. There's a LOT of luck involved.

> Even the PhD's have it really tough. No one hires PhD's, they're overqualified. Most of them end up working long hours for years as post-docs making
> $30,000.

Sadly, this is very true for many and pretty much describes my life. Sad, eh? Not surprisingly, I have been contemplating a career change for awhile now.

Any other burned-out, underpaid post-docs read Slashdot?

Oh well, back to the lab bench...

Re:What I'd major in

[Spunk]

Me too, but for the chicks rather than career interests.

Mmm, biology...

Re:What I'd major in

[Lordrashmi]

In a company that has a serious Bioinformatics department, they should have someone who has both computer experience and science experience leading it who won't only pick people with PhDs.

I work for a genetic research company and was do not have a degree of any sort. I was hired because my past work experience demonstrated I could program. I knew almost nothing about the science, but when a project is assigned to me, a scientist will explain to me the information I need to know. A programmer doesn't need to understand everything about science to program systems to support it just like a game programmer doesn't have to be a pro basketball player to program a sports game. It just takes an open mind, and willingness to listen to what your customers (in this case scientists) need.

Re:What I'd major in

[furchin]

I agree that what bioinformatics results in is very useful information, but my point is that it's not different from massive string matching. As such, if you the programmer aren't interested in string matching, you probably won't be interested in bioinformatics either.

yeah, damn biologists driving ferraris!

[Anonymous Coward]

there's a ferrari driving around downtown calgary with a plate "bioinfo"...

reminds of the "I love linux" plate someone had on a lamborghini in the late 90's, which used to be shown off at linux shows and stuff...

some people will always make money off the stuff we give for free (like genes ;)

Re:yeah, damn biologists driving ferraris!

[BWJones]

In fact, I know at least one biologist/scientist with a certain company that drives a Ferrari. Additionally, there is some other rather nice hardware in the parking lot whenever I drive by which is well deserved and hard-earned as this company is one of the hotter ones in the biotech industry with more than a little talent there.

Re:yeah, damn biologists driving ferraris!

[F2F]

see, that's because biology is harder than Computer Science -- we're simply not smart enough :)

here's an interesting undergrad project for you -- how many famouse computer scientists in the past 40 years have originally started as mathematicians.

how many as physicists?

Re:yeah, damn biologists driving ferraris!

[BWJones]

here's an interesting undergrad project for you -- how many famouse computer scientists in the past 40 years have originally started as mathematicians.

Actually, here is an interesting bit of information/history on famous mathematicians and computer scientists who have made it big. The interesting thing is that the University of Utah is the nexus that all of these guys came through.

Re:yeah, damn biologists driving ferraris!

[F2F]

funny though, none of those guys look anything like the 'i dropped out of school and look where i'm now at' type so highly touted in the pre- bubble burst era :)

interesting read nonetheless, thanx :)

Re:yeah, damn biologists driving ferraris!

[Tablizer]

there's a ferrari driving around downtown calgary with a plate "bioinfo"...

For $4.95 you can have my "petfood.com" or my "ridemir.com" plate.

Re:yeah, damn biologists driving ferraris!

[reverseengineer]

Interesting you should say that, since the foundation of molecular biology owes a lot to noted physicist Erwin Schrodinger, and his little book, "What Is Life?" Most of his speculations were incorrect- he believed that proteins, not nucleic acids, were the information carriers of the cell, for example, but as is often the case, sometimes asking the right questions can be even more important than finding answers. However, it's more than a case of a genius coming in with a bold new idea- by the late 1940s, molecular biology was an idea whose time had come. If "What is Life?" had been written in say, 1890, it would have probably been quickly forgotten- in order to make molecular biology a reality, a critical mass of organic and physical chemistry knowledge was needed, and a variety of chemical and biological techniques like X-ray diffraction, mass spec, chromatography, and cell fractionation needed to be developed.

In my opinion, the "What Is Life?" of the bioinformatics age is J. Craig Venter's whole genome shotgun sequencing method. Once again, a totally different way of doing things, and once again, from an outsider- not as much from the field of study as from every one else engaged in that field. I've had the honor of meeting Dr. Venter and listening to him lecture- he's staggeringly brilliant. He also may be the most arrogant man I've ever met. (And I've also met Stephen Wolfram.) I think often a maverick or an outsider is needed to shake things up and move things forward- either an ingenue who doesn't know the "conventional wisdom" or the hardnosed type who simply doesn't care what everyone else thinks.

Of course, once again, the new idea would have gone nowhere without thre requisite advances, this time in computing, not just in technology, but in computer science (fast algorithms so very important), and also in the development of the miracle that is the Polymerase Chain Reaction. Oh, and with regard to the title of this thread, noted biochemistry student reverseengineer is decidedly more upbeat about the idea of a bunch of "damn biologists drving Ferraris." He wants a 360 Modena, a red one.

The submitter is correct

[SteweyGriffin]

Bioinformatics is really where it's out. Just glancing around at various newspapers, it's readily apparent where the future is heading.

Is it really what we want/need as humans? I'm not sure. But I for one won't wager a guess until there's more research done in the area, so I say let's explore it more before we defame it conclusively or support it as a technological breakthrough.

Some other recent news items:
Nabda, Unesco Collaborate in Bioinformatics Training
AllAfrica.com,Africa-05 Dec 2002
... Development Agency (NABDA) and the United Nations Education Scientific and Cultural
Organisation (UNESCO), penultimate Tuesday held a two-day Bioinformatics ...

Bioinformatics ahead for Danville
Danville Register and Bee,VA-30 Nov 2002
... Developing these plants will involve both horticulture and bioinformatics and will
be one major focus of Danville's Institute for Advanced Learning and Research ...

The race to computerise biology
Economist (subscription),UK-12 Dec 2002
Welcome to the world of bioinformatics--a branch of computing concerned
with the acquisition, storage and analysis of biological data. ...

Observing Proteins And Cells In The Wild: Quantum Dots May ...
Science Daily-13 Dec 2002
... Today it is internationally renowned for research and graduate education
in the biomedical sciences, chemistry, bioinformatics and physics. ...

The future is bpu

[Anonymous Coward]

Biological processing units!

Imagine being able to create a creature which is basicly a living supercomputer! It will break teh limitations of current cpus

We already created a polio virus from scratch, and we are trying to create a organism from scratch.

(bad joke ahead)
Imagine living beowulf clusters, we could create a cell with cpu like properties, they would reproduce by splitting, and your biocomputers computing capacity would double every few minutes!

Now if each cell could perform 1 megaflop then a petaflop computer would need 1,000,000,000 cells. 2^30 is appox 1,000,000,000. So if it took 10 minutes for cells to split, then a petaflop biocomputer could be grown in about 5 hours!

Re:The future is bpu

[cybermace5]

And then it would proceed to EAT ALL LIVING THINGS before dying and leaving the planet barren, lifeless, and covered with a strange ooze.

Re:The future is bpu

[Bicoid]

Someone needs to put Blood Music down RIGHT NOW.

Re:The future is bpu

[Stephen VanDahm]

"Imagine being able to create a creature which is basicly a living supercomputer! It will break teh limitations of current cpus"

But can it find Sarah Conner?

Bioinformatics, Genomics, Proteomics

[SkewlD00d]

Proteomics will be THE next medical frontier. Maybe we will finally understand how proteins work and consequently, how living things are built. That will eventually lead to real genetic engineering and maybe an organism could be constructed from scratch.

Re:Bioinformatics, Genomics, Proteomics

[cookie_cutter]

Maybe we will finally understand how proteins work

The big problem is figuring out how proteins fold. Once we know a proteins structure, it's often trivial to figure out how it works(it's simple chemistry and physics from their).

The biggest contribution that proteomics could make would be to solve the protein folding problem.

Re:Bioinformatics, Genomics, Proteomics

[coloth]

maybe an organism could be constructed from scratch.

Of course, it wasn't created from scratch, but it may be the closest thing yet, and the most fully understood non-viral life form yet (many may have heard):

New Life Form Created

Re:Bioinformatics, Genomics, Proteomics

[spiro_killglance]

I don't think protein protien interaction is
trivial either. Lets assume that both proteins say the same shape. Take one protein
rotate it and move it around the other protein
in all possible positions and orientations the
relavent group action is R(3) tensor O(3), (the group transitions of one body in 3d, times the
group of rotations in 3d:
six dimensions. proteins are big molecules,
for each of the possible orientations and positions you need to calculate the electrical and
pauli forces between each atom on the surfaces of
one protein with those on the the surface of the
other protein to see if the there can be any chemical interaction between the two. This is a
heavy ammount of computation. But its worse than
that one protein can change the shape of another
one making the calculations even more complex.

Proteomics needs the folding problem solved.
Protein Protein interaction solved.
The inverse folding problem solved (for drug/DNA design).
Then DNA-protein interaction, and the interactions
with macro sugar molecules. Once you've done all
that and have it tabulated in vest lookup tables
for common problems
and with computation on demand for rarer problems, then you can simulate a single cell.

Re:Bioinformatics, Genomics, Proteomics

[reverseengineer]

You make an excellent point, especially since it has been discovered that in the case of many of the cell's chemical pathways, one enzyme pretty much just hands off the substrate to the next enzyme in the chain, as opposed to letting it float around in the cytosol and find the next enzyme. So yes, protein-protein interactions are going to be a very important consideration for proteomics researchers. Unfortunately, we do have to walk before we can run. The current state of things is such that we have trouble figuring out how a simple oligopeptide will fold. We must figure that out, then more complicated peptides, then multi-subunit proteins like hemoglobin or rubisco, then we can hit up the protein-protein interactions. I'd imagine this sort of computation is going to require massively parallel computing on a scale that puts today's state of the art to shame, and probably some new ideas in applying pertubation theory. All that figures to keep computational chemists nicely employed for a long time to come.

Re:Bioinformatics, Genomics, Proteomics

[Namds]

While these problems aren't trivial, all of them are already basically solved besides the protein folding problem. Protein folding won't be solved anytime soon due to computational restrains.

All of these but the protein folding problem can be viewed by ab initio techniques such as molecular dynamics. While these are very computationaly intensive calculations they are duable on super computers in a short period of time (on the order of weeks/months).

As a side note, we really don't need the folding problem solved. Crystallography has come far enough that by the time we have the computer power avalable to calculate the folding of protein, most of the protiens we know of or are of interest will already have their structures solved by crystalography.

Re:Bioengineering

[cybermace5]

What, so you can represent each possible codon as a single bit? Last I checked, you can represent 64 values with a measly six bits. (2^6 = 64).

Bioinformatics runs on Open Source

[dan dan the dna man]

So don't sit there saying its hard to get into :)

Run, don't walk to bioinformatics.org and contribute!

The first O'Reilly bioinformatics conference rocked. Shame I wont make the next one in San Diego - I get to go to Adelaide for the ISMB in June instead :)

Re:Bioinformatics runs on Open Source

[wazo2k]

Also, if you look around, there are great bioinformatics jobs that don't require specific experience/education.

I know, I've been working in bioinformatics for the last two years and I only have a major in computer science (actually, I don't, I have my last exam friday :) )

Re:Bioinformatics runs on Open Source

[glwtta]

Or go straight to bioperl.org, one of the absolute key projects for bioinformatics.

In case you are wondering, there is a biopython, biojava, biocorba, bioxml, biruby, etc. but perl is really where 90% of bioinformatics is done (simply because, in the end, all it is is text processing)

Re:Bioinformatics runs on Open Source

[sket]

you better go to Brisbane instead...otherwise you'll have to take the bus for 2000km..:) let's hope the guys there won't block slashdot as the lusers in Edmonton did. i had to login at home and use a console browser to read my favourite news site.

Re:Bioengineering

[jdiggans]

I'm not sure from whence came all your 'hences' but the sixty four combinations of ACTG only translate into 20 (some argue a few more) amino acids and start/stop signals. The system is highly redundant to lessen the impact of single-base polymorphsims (i.e. if a codon is CCA and the second C is copied as an A accidentally the same amino acid still can be produced in the end-result protein chain).

So hence ... it's not 64bit computation. It's ~23bit computation ... but all those other buzzwords are fun too. :)
-j

after overhyping the Internet

[g4dget]

now people are overhyping bio-informatics.

Yes, it's an important field. It's been an important field for decades. And it's going to continue to make steady progress, not because of, but in spite of the attention and hype, and the stupid patents and opportunism that come along with it.

Re:after overhyping the Internet

[cookie_cutter]

In spite of the attention? The attention is important to get people into the field.

The field is just starting to really contribute to society as a whole, with the first drugs developed using rational drug design being used in medical practice.

Who wants just steady progress? Most people want exponential progress in medicine, and that will only happen if the field grows.

Re:after overhyping the Internet

[BWJones]

now people are overhyping bio-informatics. Yes, it's an important field. It's been an important field for decades. And it's going to continue to make steady progress, not because of, but in spite of the attention and hype, and the stupid patents and opportunism that come along with it.

Actually, bioinformatics is just starting out and the hype has not even begun. Just wait a few years. Additionally, bioinformatics is not decades old. It really started to come together about 15 years ago, and is still an emerging field at the nexus of computer science, chemistry, genetics, physiology, anatomy, pharmacology, epidemiology, medicine, engineering, virology, microbiology, pathology, mathematics, statistics, information management and molecular biology. All of these fields are contributing rapidly exploding amounts of information and making sense of all of this information and communicating the results is what bioinformatics is all about. In fact, because of the advent of bioinformatics, new fields like pharmacogenetics, or the tailoring of drug treatments to specific individuals are being created.

For a number of reasons, the major contributor of information to the field right now is genetics. This is because genetic and protein sequences lend themselves to large-scale analysis in a much easier to encode manner and thus are easier to interpret than other types of data such as tertiary or quaternary protein structure, histology or biochemical pathways. This ease of management is because genetic sequences can be easily represented by letters which represent base pairs, and what one sees when examining a genetic sequence from DNA for example is an endless stream of the letters A, G, C, and T representing adenine, guanine, cytosine, and thiamine, the four nucleotides that make up DNA.

You are probably familiar with the relatively recent announcement of the human genome being sequenced. What has been accomplished here is that all of the A's, T's, G's and C's in the human genome have been placed in roughly the appropriate places by a consortium of both private and public research groups. What all of these base pairs mean is another set of problems that needs to be unraveled. For instance, where do genes begin and end in these sequences of letters? What regions encode proteins and which do not? These are only two of literally millions of questions that can now be asked illustrating that this rough draft is only the beginning and it is only one genome out of many that has been sequenced which will lead to a more complete understanding of organismal biology. And this is just the mammalian stuff. The real money will be on advances made in agribusiness.

Re:after overhyping the Internet

[bstadil]

I think you are wrong on both accounts. Bio-Informatics will profoundly change our lives but its Hype-factor will be less than the internet as people can not readily get a tactile "feel" for it as they can with the intenet.

Now get to my main point about the Internet being Hyped. Yes it was hyped, but rightly so.

The famous remark by Elison "The Internet changes everything" is truer today than when it was uttered a few year ago.

Go thru what you do on a daily basis and think about how this has been changed over the last few years.

All business communication (Human as well as machine to machine) is transformed. Most retail we do is transformed. Brick and Mortar retailers is relying on Internet to promote deals, We all use Google etc. to look at product information.

Science is changed forever as the Internet is now the premier reference vehichle. If a scientific paper is not available online it might as well not exist.

I could jammer on but the point is that the Hype was justified. The Economic dislocation of whealth behind the Hype was the problem, not the hype per se.

Re:after overhyping the Internet

[Tablizer]

now people are overhyping bio-informatics.
Yes, it's an important field. It's been an important field for decades. And it's going to continue to make steady progress, not because of, but in spite of the attention and hype, and the stupid patents and opportunism that come along with it.

Historically, it appears that investors (and career planners) are not very good at matching the "ramp-up curve" of new tech. The web and HTTP are changing the way business works, but the pace may be more gradule, and in different ways, than investors anticipated.

Technologies like automobiles and telephony did become big industries, but investors over-guessed the first few years and under-guessed a decade or two down the road.

History suggests a gradual, but substantial ramp up in progress for promising new technologies. I am going to wait for some kind of mini-crash about 5 years from now before I invest in biotech stocks. I suspect that the payoff will be slower than impatient investors want, and then it will be a fad to dump biotech stocks after a big biz magazine has a headline like, "Biotech repeatedly dispoints investors". That is the point to jump in.

But, you never know for sure. Predicting the future is a tough biz. The internet has shown that the proliferation of a given tool does not necessarily guarentee profits. It may become a simple prerequisite for survival rather than a profit machine.

Re:after overhyping the Internet

[glwtta]

is still an emerging field at the nexus of computer science, chemistry, genetics, physiology, anatomy, pharmacology, epidemiology, medicine, engineering, virology, microbiology, pathology, mathematics, statistics, information management and molecular biology

Actually, it's a biology background and a bit of perl. (I personally do the perl part).

Re:after overhyping the Internet

[g4dget]

Additionally, bioinformatics is not decades old.

Sure, it is: DNA sequence databases, genetic sequence analysis, simulations of regulatory networks, simulations of biochemical networks, 3D structural analysis of proteins, etc. were already in wide use in the early 1980's, and a lot of that goes back to the 1960's.

You are probably familiar with the relatively recent announcement of the human genome being sequenced.

But you are evidently not familiar with the history of biochemistry or molecular biology. Bioinformatics didn't start with the attempts to sequence of the human genome.

These are only two of literally millions of questions that can now be asked illustrating that this rough draft is only the beginning and it is only one genome out of many that has been sequenced

These are the same questions that people were asking before the sequencing of the human genome, and many people still believe that what the human genome project accomplished was not the most cost effective approach to answering them. And it is not the case anyway that the human genome project has delivered an end-to-end sequence of the human genome.

Re:after overhyping the Internet

[g4dget]

Bio-Informatics will profoundly change our lives but its Hype-factor will be less than the internet as people can not readily get a tactile "feel" for it as they can with the intenet.

Molecular biology will profoundly change our lives. Bioinformatics is only one of many tools of molecular biology.

Now get to my main point about the Internet being Hyped. Yes it was hyped, but rightly so. The famous remark by Elison "The Internet changes everything" is truer today than when it was uttered a few year ago.

The Internet did change everything, but, like bioinformatics, it was created quietly in the 1960's and grew steadily for several decades afterwards. The hype and feeding frenzy in the late 1990's hurt the further development of the Internet and thrust us into a recession.

So, optimism about both bioinformatics and the Internet is justified, but what commercial applications need is steady, continued investment, not a huge influx of people and money all at once.

Re:after overhyping the Internet

[g4dget]

The field is just starting to really contribute to society as a whole, with the first drugs developed using rational drug design being used in medical practice

Rational drug design has been around for a long time, and that is not the "first drug" designed rationally. Of course, decades ago, people couldn't do rational drug design by the kinds of large scale simulations we do today, but the principles of rational drug design were still the same.

Who wants just steady progress? Most people want exponential progress in medicine, and that will only happen if the field grows.

Using the term "exponential" similar to "rapid" generally suggest someone who has no idea about science; a steady 3% annual growth rate is "exponential growth", just like an an annual doubling.

In any case, the problem with the hype and bubble approach is that after a few years of spectacular growth and job opportunities, you get many years of spectacular failure and disappointment.

Re:after overhyping the Internet

[BWJones]

DNA sequence databases, genetic sequence analysis, simulations of regulatory networks, simulations of biochemical networks, 3D structural analysis of proteins, etc. were already in wide use in the early 1980's, and a lot of that goes back to the 1960's.

Of course they were. However back in the 60's, 70's or 80's or even 90's I could not perform quick sequence searches of various species, get a 3-D appreciation of a particular protein and then design and submit an antibody or drug for that sequence in an afternoon. In order to do this, I need databases that have been populated with information developed over decades and tools to mine those databases. That is some of what bioinformatics is about.

But you are evidently not familiar with the history of biochemistry or molecular biology. Bioinformatics didn't start with the attempts to sequence of the human genome.

A bit snippy are we? Actually, I am quite familliar with the history of biochemistry and molecular biology and I never said that bioinformatics began with sequencing the human genome. Rather I was using it as an example as it was widely publicized and well known. Please re-read my previous comment.

Re:after overhyping the Internet

[g4dget]

Of course they were. However back in the 60's, 70's or 80's or even 90's I could not perform quick sequence searches of various species, get a 3-D appreciation of a particular protein and then design and submit an antibody or drug for that sequence in an afternoon. In order to do this, I need databases that have been populated with information developed over decades and tools to mine those databases. That is some of what bioinformatics is about.

Since you admit that the technologies and software were already available decades ago, what is your point? Are you saying that bioinformatics is about being able to do something in an afternoon, as opposed to having to run it on a mainframe over a few days?

I never said that bioinformatics began with sequencing the human genome. Rather I was using it as an example as it was widely publicized and well known. Please re-read my previous comment.

You wrote:

Actually, bioinformatics is just starting out and the hype has not even begun.

A bit snippy are we?

No, not at all. It is simply evident from your post that you are a bit taken with the hype and a bit unfamiliar with the history of molecular biology, even if you talk at length and repeatedly about "T's, G's, C's, and A's".

Re:after overhyping the Internet

[bstadil]

Both your points are well taken, notably the need for investments to be steady, long-term focused and balanced. Cheers

Re:after overhyping the Internet

[BWJones]

Since you admit that the technologies and software were already available decades ago, what is your point? Are you saying that bioinformatics is about being able to do something in an afternoon, as opposed to having to run it on a mainframe over a few days?

Sigh.......No......I did not admit that the software was already available in the 60's, because it was not. Bioinformatics is not about individual technologies such as protein crystalization or running gels or finding and quantifying enzymatic activity. Rather it is the overall integration of the knowledge of the individual technologies, and the management and dissemination of the data. This requires databases and tools with which to mine the data present in databases. Yes, we did have databases decades ago, and yes, there was some analysis of those data. However, large scale integration of many different types of data from many different disciplines and techniques was not accomplished until relatively recently.

No, not at all. It is simply evident from your post that you are a bit taken with the hype and a bit unfamiliar with the history of molecular biology, even if you talk at length and repeatedly about "T's, G's, C's, and A's".

What is exciting is that much of this work can now be accomplished on the desktop and for very little money compared with what was possible only a few years ago. As for your griping about familiarity in molecular biology, my original post was simply to inform the general Slashdot community and not get into a pissing match with you or anyone else with an axe to grind or a thread to troll. As for overall competence in science, I would rather rely on my scientific publications, record of work and reputation with colleagues to demonstrate my knowledge and capability as opposed to taking any more time with you shouting into the wind on Slashdot.

Best,

Re:after overhyping the Internet

[g4dget]

I don't question your scientific qualifications (which I don't know anything about), I question you judgement. I have been through several cycles of hype-and-bust in different academic disciplines. It derails scientific careers, slows down research, locks up scientific results in corporate vaults, and is generally destructive. It also attracts opportunists to fields in which they have no interest or skill.

Progress in bioinformatics has been steady over decades. Many of the capabilities we have now are based on methods and ideas developed a long time ago, although they have much more impact now with powerful PCs.

What we need is steady, continued public research funding in the application of computers to biological systems. What you advocate, and what follows from hyping up a field, rapid expansion and large new private investments, is destructive in the long run.

Re:after overhyping the Internet

[BWJones]

What we need is steady, continued public research funding in the application of computers to biological systems.

Absolutely.

What you advocate, and what follows from hyping up a field, rapid expansion and large new private investments, is destructive in the long run.

I would agree with you here, but I don't think that public attention to an issue is always necessarily a bad thing. It certainly is bad for the science if the "spin" is put on issues to make them more than the science supports. Additionally, I would agree that hyping of any discipline beyond the public expectations can bring difficulty to any field (the U.S. space program, the Internet, the cardiac transplant field, the whole retinal transplant field right now, etc...et ...etc...) and I would never advocate spin beyond what the science currently supports.

Competent columnists...

[pVoid]

And empirical estimates are being replaced by mathematical exactness

I'm not dissing the article completely yet (as I haven't finished reading it, and don't know if I'm completely interested), but I find it wonderful how ignorance among press still prevails.

There is no science, (apart from Math itself - which I consider more of an Art) that has mathematical exactness in it. The word science comes from the latin root of scientia, and means knowledge. Sciences are disciplines where as much knowledge about the existing (thus empirical) world is gathered as possible, and models are generated based on this data.

Mathematics on the other hand, derives from Axioms, and Logic. Both of which aren't derived from the empirical world. And I say it's much more akin to Art because it is a skill that you develop to be a mathematician: you forge out of simpleness new more complex theorems. You are 'creating' them... (in science, you are looking for them).

To make a long story short, there is no such thing as a mathematically exact science.

Re:Competent columnists...

[DannyO152]

I wouldn't be so confident that the base axioms are so separate from empirical observation. Euclidean axioms with regards to the sum of angles of triangles and parallel lines were motivated from the experience of the Greeks, Babylonians and Egyptians. Certainly since the 19th century, when mathematics became concerned with completeness of axiomatic systems, the results of logic applied to a system of axioms did not require a naturally observed analogue for validation.

Meanwhile, some 17th and 18th century advances in mathematics occurred because observations produced results that required new mathematic techniques to explain. It seems to me that the calculus was initally accepted because it worked for the laws of motion, rather than that there was a unassailable proof that the limit of (x2-x1) as x2 approaches x1 can be divided into the numerator and then once that pesky denominator is gone, we can let x2 = x1. And therefore if f(x) = x^2, f'(x) = 2x

Mathematics is a discipline which may mean it's a craft. I don't think it is an art, because Art allows for idiosyncratic solutions which makes one artist distinguishable from other artists, even when approaching the same subject. As for forging and creating threorems, Godel and Church developed a way of enumerating obtainable results from a set of axioms and this implies that all the provable theorems exist, but like the next prime number, we haven't gotten there yet. Obviously the next theorem is not determined by giving a computation machine a number, such as 42, and waiting for it to expand into a theorem. Intuition and sweat are required, but once applied, the proof must withstand the scrutiny of the learned. The results of an Artist's intuition are accepted without requirements of proof, and sometimes that acceptance or rejection is so subjective as to be arbitrary.

Mathematics is the language of science, and has advanced when science required new "words and sentences" to describe empirical results. By the way, I do completely agree with your comment about the babble of the quoted phrase. All observation have errors of precision -- we repeat experiments and use statistics and probability to interpolate a demonstration that the underlying hypothesis, often described mathematically, is valid. An interesting book I just finished which touches on the issue, and how it changed Philosophy, Education, and Law is "The American Metaphysical Club" by Louis Menand (Farrar, Straus & Giroux).

Viewpoints

[VoidEngineer]

Hmmm. Seems like it would be easier to say that everything which has been done in computer science has 'been done', whereas everything that hasn't been done in computer science, 'hasn't'.

Seriously, though, you may be mis-categorizing your subjects. Look at computers as computational entities, rather than disk drives, monitors, and so forth. In that case, an optical computer or a biocomputer operates on many of the same systems priciple as a 'digital computer', and there is therefore much to be done in the field of computer science.

Absolutely. Optical computing is getting some great advances in Holographic Video at the MIT Spatial Imaging Group. And chemical computing is advancing nicely in Carbohydrate Chips at the University of Chicago.

For my money, I'd bet on optical video cubes, 3D television, and biochips in the future... which are all applications of computer technology. Remember, 'computer' use to refer to the job title of a person.

For my money, I think that the future has got SnowCrash, Cryptonomicon, Neuromancer, Count Zero, Mona Lisa Overdrive, and Johnny Mnemonic written all over it (and maybe a bit of Jurassic Park.

Re:Viewpoints

[Bicoid]

For my money, I think that the future has got SnowCrash, Cryptonomicon, Neuromancer, Count Zero, Mona Lisa Overdrive, and Johnny Mnemonic written all over it (and maybe a bit of Jurassic Park.

Sure, the future has Neuromancer written all over it. But it also has Schismatrix written all over it as well. We're likely to see biological modification (including genetic alteration) in the future. Even if the majority finds gene-tampering taboo, there WILL be groups that will do it.

The truth is, we're likely to see a combination of the two in the future. Yes, we're likely to see a computer-saturated world within the next few years (if you don't consider it computer-saturated right now), but that doesn't mean we won't see designed genetics as well. The two work together. It's not an either/or issue.

Re:Viewpoints

[Lemmy Caution]

I think that the era of the general accessibility of computing is coming to an end, if it isn't over already. Biotechnology and bioscience are difficult, complex, highly specialized and incompletely understood fields - the number of people who can do anything resembling "hacking" is miniscule. Biotechnology will reach people almost entirely as products and services, not as a realm of activity in which they can directly participate.

I think computing is going to move in that direction, as well. Libraries are much too complex and interdependent to be understood by the programmers who use them. We are more likely to manipulate large libraries and pre-programmed objects, working in increasingly visual environments with only a few specialists tweaking specific low-level libraries (and those specialists not necessarily understanding each other's work). I think that there already is a sort of nostalgia for the epoch of computing when any dedicated amateur could pretty much comprehend what was going on in their machine.

bad idea

[g4dget]

Don't pick what is fashionable, pick what you are interested in. The biological sciences are very different from computer science and engineering. Chances are that if you weren't interested in the biological sciences when you picked your major, you still wouldn't be today.

What you can do, however, is apply computer science and engineering skills to biological problems: work as a developer or engineer for a biotech company or lab.

Revolution

[Raiford]

This is one field that is under an dramatic revolution. The computer has done wonders in enhancing physical sciences such as physics and chemistry as well as engineering. However there is no questioning the impact this technology has had in this field. Just think the structure of DNA is post WWII science (still relatively young by comparison with physical theories). There is some much more to learn !

Re:Bioengineering

[yerricde]

Last I checked, you can represent 64 values with a measly six bits. (2^6 = 64).

But if you're searching for some pattern, representing each possible alternate codon as a separate bit has its applications.

Re:Bioengineering

[VoidEngineer]

What, so you can represent each possible codon as a single bit? Last I checked, you can represent 64 values with a measly six bits. (2^6 = 64).

That's fine, but the point is that one can make sort of a 'virtual gene' or 'genetic codon probability template', which happens to be optimized for a 64 bit computer, utilizing genetic algorithms. I agree with you that you can represent 64 values with six bits, which isn't exactly what I was refering to. The idea is to map each possible codon onto a bit of the packet which is going to analyzed by the processor. Probably, one will want to create a big-Indian or little-Indian (endian, whatever) ordering scheme.

For example:
AAA = 10000000000000000000000000000000000000000000000000 00000000000000
AAC = 01000000000000000000000000000000000000000000000000 00000000000000
AAG = 00100000000000000000000000000000000000000000000000 00000000000000
AAT = 00010000000000000000000000000000000000000000000000 00000000000000
etc.
and so forth. Therefore,
11110000000000000000000000000000000000000000000000 00000000000000


would equal ( Protein | AAA, AAC, AAG, AAT).

Great for optimizing a mainframe for crunching through a combinatorial space of codons, searching for proteins.

My $.02 on biological sciences

[t0rnt0pieces]

I graduated from Rutgers U. (decent NJ state school) in May 2000 with a bachelor's in biology. Back when I decided to major in bio, I really enjoyed studying the field, and (high school) teachers were telling me that molecular biology was the fastest growing job sector. So by the time senior year came around I began looking for a "real" job. I checked all the papers, company websites, monster, etc. and realized that there really are NO good jobs out there for biologists. There are a few bioinformatics jobs, but guess what, you need a CS degree for them, not biology. I ended up getting a lousy job as a lab technician paying around $14 an hour (which took several months to find, btw, and this was when the economy was booming), and I'm extremely unhappy. I've completely lost any love I had for this field. Say what you want about money not being the most important thing, but when you can't afford to do anything you want, your life gets miserable very fast, I don't care how great your job is. Biology jobs typically have zero mobility and are extremely underpaid. You think there's a flood of people graduating from CS? Biology is far worse. My CS classes have about 20-30 people in them. My bio classes had hundreds.

Just recently I decided to go back to Rutgers for a second bachelor's in computer science. Not only are the job prospects better and higher-paying (even considering the dot-com collapse), I've always enjoyed working with computers and my grades are actually far better (although I was never a bad student). I'm actually having fun in my CS classes, I never had fun in bio. The classes were more like a chore. Another horrible thing about biology is that you really don't learn anything practical in college, you just memorize facts. At least with computer science you learn many useful tools to make you a competant programmer, which is actually a marketable skill. Nothing about biology is marketable. I highly discourage anyone from majoring in Bio unless they seriously, seriously love it and intend to get a PhD and devote their lives to research without regards to trivial things like money and job prospects.

Re:My $.02 on biological sciences

[t0rnt0pieces]

You aren't going to go far with just a BS or BA in biology, particularly molecular biology (speaking as a molecular biologist). There are plenty of job opportunities in biotech for PhDs, though...

As far as I know there aren't many jobs for PhD's either. I work with someone who has a PhD in Chemistry (ok, it's not bio, but chem's probably a better field than bio), and he's a lab tech like me!! Besides, why the fsck should anyone have to get a PhD just for the chance to make a livable wage in their field? If that's the case, I'd say it's a lousy field.

Re:My $.02 on biological sciences

[Tablizer]

You think there's a flood of people graduating from CS? Biology is far worse. My CS classes have about 20-30 people in them. My bio classes had hundreds.

That is because they all want to know how to make LSD and Extacy in their dorm, not a bio career :-)

Re:My $.02 on biological sciences

[freeweed]

Wow, your story parallels mine eerily. I graduated in 1996 with my B.Sc., microbiology major. Jobs that paid enough to live off were just about non-existent (this is what happens when several hundred people graduate with your degree EVERY YEAR in a city of half a million people).

Now, I'm 3/4 of the way through a CS degree, and enjoying school like I never have before. Sure, part of it is a little more maturity/experience, but I find now that I'm in something that I ENJOY, I couldn't care less if I'm learning about NP-Completeness today, or Java Beans tomorrow. Regardless, I'm about to enter a co-op work term that will pay $18/hr - and this is one hell of a lot above average salary around these parts (think closer to $12), even for people with university degrees (maybe $15 if you're lucky).

Good money, really fun work, this is the life. It's certainly NOT what I experienced in the bio field. Ironically enough though, I'll be graduating with my CS degree just as 'bioinformatics' (they've changed the buzzword for this a few times since I was in high school) is supposed to hit the big time. Would be kinda neat to have that microbio degree be worth somethign after all :)

Re:My $.02 on biological sciences

[efatapo]

Well, as sorry I am that you had a bad experience as a Biology major, I thoroughly enjoyed it. I got my undergraduate degree in Biology from Hillsdale College in Michigan and minored in Chemistry. If you don't enjoy your major, change it. Seems pretty simple to me.

And the fact that there is a large number of people in Biology classes I would hypothesize is related to the number of people who try to get into Medical School. And veternary school and graduate school in biomedical sciences which is a huge and growing field. Whereas I don't think CS majors have that many options.

I am currently getting a PhD in Pharmacology/Biochemistry but just wanted to offer a counter opinion as someone who LOVED Organic Chemistry and Biochemistry and Cell Biology and Developmental Biology. Biology and Chemistry are absolutely fascinating. If you don't enjoy them, don't major in them...just like if you don't enjoy CS....don't major in it. I think that's a better message to send then "I highly discourage anyone from majoring in Bio".

I assume you didn't get a teaching certificate or take any business classes where you could have easily done hospital finances. Or did you look at Government jobs as an inspector of the numerous things the government regulates. Or look to OSHA or some college/company similar organization for training on safety issues. All can be based on a biology degree. Just a couple counter thoughts from a person who loves biomedical science!

Re:My $.02 on biological sciences

[Black_Logic]

So by the time senior year came around I began looking for a "real" job. I checked all the papers, company websites, monster, etc. and realized that there really are NO good jobs out there for biologists.

May be a little OT, but, I think there is a moral to this quote. When deciding on a career path, take some of the steps you would to land a job after you get a degree to get a better feel for the field.

Re:My $.02 on biological sciences

[Bald Wookie]

Well, as sorry I am that you had a bad experience as a Biology major, I thoroughly enjoyed it. I got my undergraduate degree in Biology from Hillsdale College in Michigan and minored in Chemistry. If you don't enjoy your major, change it. Seems pretty simple to me.

No, the correct advice is this:

Unless you plan on going to graduate school for a long time, do not major in the sciences. Without a PhD you're a peon. There's no respect and little room for advancement. Everyone knows that you were pre-med and got one too many B's to continue.

I've got a BS in Chem and missed a Bio minor by a class or two. I loved organic chem, mollecular bio and all of those classes. When I graduated, I switched to the IT field because it paid twice what most science jobs paid.

Who wants to go through all of the rigor of a science major just to get a job as some government inspector? Sorry, I won't ever suggest that anyone sweat through o-chem to become the OSHA guy. If you truly love the field you won't be satisfied with those kind of jobs anyway. Making sure some binder has all of the MSDS sheets doesn't have anything to do with my love for science.

Here's my warning to all high school kids and undergraduates. You won't hear this advice from the admissions department or your advisors.

Do not major in a science. If you do, you should plan on doing one of two things:

1. Get a MD or PhD
2. Double-major in business, or maybe CS

Stay away from option number 2 unless you specifically want to combine the two fields or option 1 doesn't work out. College is a great time in your life. A science degree will take more of your time than most other disciplines. If you don't plan on reaping the benefits, don't put in the work.

There's something to be said for doing what you love. However, what good is it to study what you love in college if you can't practice it once you're out?

Re:My $.02 on biological sciences

[fain0v]

I HAVE a degree in microbiology and computer science. Keep in mind that you need experience in both to get a bioinformatics job. There is no such thing as an entry level bioinformatics job.

coin analogy?

[Tablizer]

More commonly known as "gene chips", microarrays are to the genetic revolution of today what microprocessors were to the computer revolution a quarter of a century ago. They turn the once arduous task of screening genetic information into an automatic routine that exploits the tendency for the molecule that carries the template for making the protein, messenger-ribonucleic acid (m-RNA), to bind to the DNA that produces it. Gene chips contain thousands of probes, each imbued with a different nucleic acid from known (and unknown) genes to bind with m-RNA. The resulting bonds fluoresce under different colours of laser light, showing which genes are present. Microarrays measure the incidence of genes (leading to the gene "sequence") and their abundance (the "expression").

The analogy that comes to mind is a coin sorter. Is this an accurate analogy? It also appears that there is not necessarily an existing "slot" for many genes, like what happens if you get a coin from a country that your design did not include. You don't know where it will end up.

Re:coin analogy?

[cookie_cutter]

there is not necessarily an existing "slot" for many genes

Correct, in which case you won't detect the unknown gene.

However, the chance that an unaccounted for gene will falsely trigger another detector "slot" is VERY unlikely; each detector is exceptionally specific for what it is supposed to detect.

Re:coin analogy?

[Tablizer]

Correct, in which case you won't detect the unknown gene. However, the chance that an unaccounted for gene will falsely trigger another detector "slot" is VERY unlikely; each detector is exceptionally specific for what it is supposed to detect.

Okay. How do you know a gene is "missing" then? Is it that you account for the frequent ones, and if all those slots are empty, you then know it is a rare one? IOW, how much does mutual-exclusive deduction play a role?

Re:coin analogy?

[cookie_cutter]

How do you know a gene is "missing" then?

You don't. Gene arrays can only detect genes which they are built to detect. They cannot determine whether unaccounted for genes are or are not there

how much does mutual-exclusive deduction play a role?

Not much, unless you know that their is some gene in your sample that you expect to be detected and want to see if it matches one of the genes on the detector array.

Re:Bioengineering

[VoidEngineer]

Thank you. =)

I agree that the problem can be implemented with a 23bit model. I believe that model is valid in many cases. However, I see the 64 bit model as being more accurate and representative of the underlying mathematics of genetics. Moreover, one can do the calculations with a 64bit model, and utilize a 23bit model as a check... if the 64bit model runs correctly, the results should be reducible to a 23bit explanation, and can be checked via another algorithm.

Also, I believe that the 64bit model can model a bunch of stuff that the 23bit model leaves out, including mutations, loops, reverses, and so forth. (i.e. aaa.aaa.CCC.ggg.ggg.ttt.ttt becomes aaa.aaC.CCg.ggg.ggt.ttt.ttX) I forget what that's called, but you get the idea. Deletion! Ah. Deletion, mutation, etc.

Anyhow, the 64bit model is really slick mathematically, and one can do really crazy cool masking and information analysis on the matrices which are formed.

Re:Bioengineering

[shirai]

I kind of hope I don't get modded down for this as I am totally serious: This is one of those posts where I didn't know whether to mod it "Insightful" or "Funny." Perhaps we need the new mod: "Lost"

Security

[akuma(x86)]

Perhaps some of these bioinformation engineers should spend a little time on security. I tried to go to the website of one of the companies referenced in the Economist article and got a defaced website:

Wow!

[c0dedude]

'Bioinformatics is not for the faint of heart...'"
... They're not kidding! Even the name is hard to spell!

As a bioinformatics scientist...

[smoondog]

Bioinformatics is a fun interesting field. I worry however, that it may be a little overhyped. People who are interested in bioinformatics need to realize it is a very (albeit cool) specialized field. There aren't going to be a million more bioinformatics researchers in the world. The demand for these researchers just isn't there (and won't be there in the near future). For example, a search on hotjobs reveals only 51 listings for the keyword bioinformatics and nearly 900 listings for programmer.

That said, bioinformatics is exciting. If a computer savy person is interested in getting into it, they should intern or work for a researcher/professor on a research project. You might be surprised, however, when you find that working as a programmer elsewhere pays 2-4 times more.

-Sean (sdm@stanford.edu)

Re:As a bioinformatics scientist...

[Anonymous Coward]

The demand for these researchers just isn't there (and won't be there in the near future). For example, a search on hotjobs reveals only 51 listings for the keyword bioinformatics and nearly 900 listings for programmer.

And strangely none of the @#%! 900 reply. Go figure. I think they are just going through the motions for paper-work reasons. Probably H-1B permits require an ad be published or something, and Hotjobs was cheaper than the paper.

Nobody wants goddamned programmers these days, except for the very top resume liars, you see? Thus, we are all digging around for ideas to get a paycheck doing geek stuff. Your observations are not very encouraging.

Using your info, it basically boils down to:

bio < programming

programming = 0.00001

Therefore: bio < 0.00001

(The unit of measure I am using here has something to do with snowballs and Hell.)

Don�t forget Chemoinformatics

[Wdi]

While there is a real hype about bioinformatics, do not forget its sister discipline, chemoinformatics. Chemoinformatics is more concerned about handling chemical structures rather than genomes (but the boundaries are not that clear, and companies such as Accelrys 8cited in the article] are more a chemoinformatics than a bioinformatics company).

An interesting overview about CI can be found at Nature.

Still, you need dedication for this job: A Ph.D. in chemistry plus solid computer science knowledge is still the norm. But those few who qualify are really sought after.

Disclosure: I am the Director of Chemoinformatics at start-up ChemCodes (www.chemcodes.com), so I know what I am talking about.

Re:Bioengineering

[GimmeFuel]

Same here. I just got my mod points message, and starting browsing this article to see if there was anything I should mod. However, instead of "Lost", what I think what we really need is "+1 Completely confusing to the mod but the poster sounds like he knows what he's talking about"

Re:Bioengineering

[VoidEngineer]

Nah. I'm not exactly 'lost'. I did live in the hometown of Gregor Mendel for a year during high school, where I studied mendelian genetics (Received a scholarship, via the 1994 congressional information act, and went with the Youth For Understand program). I also studied genetics at the University of Chicago's Department of Ecology and Evolution. Anyhow, I happen to know alot about genetics, actually... to the point that I'm making comments based on 'insider information' much like many people make 'inside jokes' which don't make sense to other people.

I would suggest the new mod: "Inside Info"

What we all want to know

[Tablizer]

Bioinformatics is a fun interesting field. I worry however, that it may be a little overhyped.

Okay, being burned by past bubbles, how can I *this time* around make money from the poppage of future bubbles? (No, "stock puts" are too expensive for me.)

Postdocs Jobs in bioinformatics and proteomics

[goombah99]

Want a high paying job in bioinformatics? there are several high paying postdoctoral positions ($59000 to $75000 starting salary) available at Los Alamos National Laboratory.

goto www.lanl.gov and click on the 'jobs' tab then the postdoc link.

here is one example:

Summary: Postdoctoral Positions in Protein Bioinformatics and Structural Genomics: The Bioscience Division (B-2 Group) is seeking 2-3 highly motivated researchers for immediate openings to work with our interdisciplinary team of Bioinformatics and Structural Biology. Research activities will focus on the development and application of methods in Functional and Structural Genomics, including: 1) inference of function in proteins based upon structural and sequence information; 2) prediction of protein structure, protein binding, ligands, and active sites using both ab initio approaches and experimental information; 3) identification of signatures of pathenogenosis; 4) annotation and analysis of selected genomes; and 5) creation and curation of annotated protein databases.

Required Skills:Experience in at least 2 of the following areas is required (more than two areas of experience is highly desirable):

• Protein structure modeling or protein-ligand analysis or other related modeling
• Background in molecular biology, or microbial pathogenesis, or related fields
• Experience with the common sequence analysis tools for Blast search, sequence alignment, phylogenetic analysis, etc.
• Drug design, or protein design or protein structure predictions or docking
• Functional annotation of putative genes based on literature analysis
• Curation of biological databases and web programming

Desired Skills:Knowledge of one computer programming language (e.g., Perl, Python, FORTRAN, C++). Use of common molecular graphics tools such as Pymol, Xtal. Research in genomic sequence analysis or protein structure. Familiarity with SQL databases, unix, and XML is useful. Education:A Ph.D completed within the last 5 years or soon to be completed is required. Notes to Applicants:Starting salaries range from $59,300 to $67,300. For further technical information about the position and the project, contact Charlie Strauss at cems@lanl.gov (505-665-5838), or Murray Wolinsky at murray@lanl.gov (505-665-0952).
Candidates may be considered for a Director's Fellowship and outstanding candidates may be considered for the prestigious J. Robert Oppenheimer, Richard P. Feynman or Frederick Reines Fellowships. Please see Special Postdoctoral Fellowships for further details.

For general information refer to the Postdoctoral Program page.

Best sources of bioinfo ... for the curious ...

[Anonymous Coward]

http://www.ncbi.nlm.nih.gov/

and

http://genome.ucsc.edu/

Re:Bioengineering

[VoidEngineer]

Ah, just what I need: another sarcastic anonymous coward.

Sorry for sharing knowledge. I'll keep my mouth shut and write closed source code in the future.

Re:64 combinations != 64 bit

[VoidEngineer]

Ummm. That's exactly my reasoning. 26bit computing is perfect for analyzing language, but only as it is written in the Roman Alphabet. Other alphabets, such as Cyrillic, would need other platforms.

And exactly my point that they are both related to molecular biology. But seriously... James Watson and crew, who theorized the double helix structure of nucleic acids back in the 50s, were part of a growing trend of molecular biologists who were investigating stereoscopic visualization of chiral molecules. And you are correct that visualization techniques and databases don't necessarily have anything to do with each other. Genetics, however, encodes what is fundamentally a chiral molecule and is therefore fundamentally encoding stereoscopic data. So, there is a connection, exactly at the molecular biology level.

Re:Bioengineering

[reverseengineer]

I'm not certain I follow the reasoning as to why 64-bit computing is ideal for genomics. I mean, it's generally going to be faster and more efficient than 32-bit computing, but that really has little to do with codons. I don't know why you would need to assign a 64-bit number to an element in a set of 64 elements- as other posters have pointed out, you'd only need a 6-bit set of numbers to label 64 things. Besides, saying there are 64 codons is a tad naive, since it doesn't account for things like post-translational modifications to amino acids and nonstandard tRNA anticodons. I hope no one has tried to study the translation of something like collagen (full of modified amino acid hydroxyproline) thinking that the codons were the last word in the formation of the protein. I don't see why 64 bits is optimal as far as the crunching is concerned- are you saying that a 128-bit computer program would for some reason not be as suitable to the task?

Re:Bioengineering

[VoidEngineer]

Oh, I think that 128bit would be a good platform, as well as any other multiple of 64, especially 256bit computing. Although, due to chunking, I would not attempt to do serious genetics work on a 32bit computer.

And please don't think that I'm trying to give the last word on this subject. I'm merely trying to point an optimal method for number crunching. There are certainly many optimal algorithms. I suspect that each algorithm will ultimately produce different results, so it's important to consider methods to be used before going out and spending alot of time and money coding a project.

And, as you said, 128bit could do the task. 64bit chips are more available on the open market at the present time. Now, a 256bit processor, with an 8bit coprocessor, could do some amazing work in a number of applications which need to compute matrices.

I'm not trying to say that what I've suggested is the last word on this subject. Far from it. A great book on the subject is:

Genetic Algorithms + Data Structures = Evolution Programs. Zbigniew Michalewicz. 1996

Re:Bioengineering

[RackinFrackin]

I don't know why you would need to assign a 64-bit number to an element in a set of 64 elements- as other posters have pointed out, you'd only need a 6-bit set of numbers to label 64 things.

Representing each codon as a 64-bit word with a single nonzero bit would speed some operations up. Using this representation, permutation group operations can be programmed very efficiently using only loops and bitwise operations. Trying all possible permutations of a list of codons could be done really fast by blazing through them in Gray order. In this case, 64 bits is the minimum wordlength that you'd want, so that you can fit a whole codon into the CPU. As usual, the larger the wordlength, the better, since you could deal with more codons at a time.

Re:Genetic algorithms != bioinformatics

[VoidEngineer]

except if you happened to apply a genetic algorithm to a bioinformatics problem

Well, yes. That is actually exactly what I was implying.

I happen to think that the book is quite good, and I have read it, and I do know what it's about. In fact, I took a class in which it was one of my primary reading sources. The class was entitled 'Cultural Evolution and the Dimensions of Globalization'.

Now then, there are many different ways to skin a cat. There are also many different ways to write an algorithm. I am merely saying that this book is a good source for learning more about optimisation problems, and how to code evolution programs, utilizing genetic algorithms. When you finish the book, you will understand that 64bit computing is an ideal (although not necessarily perfect) platform for bioinformatics and genetics work. Yes, I agree that it is not the only platform, but it is an ideal platform because a 'genetic data chromosome' can easily be written for all of the codons utilizing a 64bit vector.

Got a BS+ in Bio? Get a JD!

[crc32]

Hey, I have a BS in Bio, and I am about to complete my law degree. I will soon be practicing patent law, which has always been a secure field. I know how bad the general /. community feels about patents, but they are a fact of life, and I for one think they are very important for technological advancement. (Are scientists worse in the EU? No? Then why is US consistantly leader in innovation: liberal patent system)

Why slave away at a lab bench for $30k when you can make $100k starting as a patent lawyer?

The moral of this story is - get a PhD

[citanon]

If you're really serious about doing any type of biology or chemistry, a PhD is a requirement. Physics people have more in the way of engineering jobs at the MS level, but also need PhD to do serious research.

I'm an MSCE ! where do I apply?

[guest12]

great opportunity for all of us MSCSE's.

Re:Uh... an algorithms textbook in a politics cour

[VoidEngineer]

Yes, actually. The point of the reading was to gain a greater understanding of how to optimize traveling salesman problems for distribution of global resources. We weren't just talking about cultural evolution and the dimensions of globalization to make us feel better... Rather, we were going through the mathematics of how to solve the traveling salesman problem and calculate memetic distribution amongst society. Topics included:

evolutionary genetics (genetics, memetics, bioinformatics, change management)
epidemiology (vector theory, networks, viruses, propogation, transmition)
demography (demographics, statistics, data mining, forecasting)
economics (markets, networks, advertising, buy/sell functions)
communications (telcom, network programming, routers, collaboration, push/pull)
science and technology (mainframes, personal computers, networks, design)
history (memory structures, databases, file systems)
political science (US Code, social programming)

Anyhow, those were most of the topics covered. It was a graduate level sequence, and I worked in a network programming laboratory at the National Opinion Research Center while I was taking the course. The other reading for that class I used included Knuth's The Art of Programming, Hull's The Structure of Scientific Revolutions, Plato's Republic, and all of Wimsatt's personal publications.

Oh, and the course was taught by four people: a memetic evolutionist, a linguist, a mathematician, and a computer programmer. We would use an algorithmic template (the genetic algorithm) and create an instance and map that algorithm onto each of the above mentioned problems and discuss the pros/cons regarding implementation. There wasn't much purpose of taking the class if one didn't know how to optimize an algorithm.

PhD positions in bioinformatics?

[lude]

I have recently finished my master's degree in biomedical sciences, my thesis and traineeships being all about bioinformatics.

Now I am wondering what country and especially which institutes offer the best atmosphere to do a bioinformatics PhD well. Does anybody have advice?

Postdoc required

[heroine]

The only problem is biology is something everyone wants to do. It's what all the highest rated TV shows are about. The people who do it are celebrities on not just geek websites but real news. You have to spend a long time in school and a lot of money to get an entry level position anywhere in it because all the fellowships are taken by celebrities. By the way, biology is not a good degree to go into bioinformatics. Chemistry is where you should be.