A farewell to bioinformatics (2012) • Hacker News Board

Re: A farewell to bioinformatics (2012)

#21

Post by michaelhoffman » Sun, Jan 27, 2013, 5:04 AM UTC

zerohp wrote: post #6

> the software is written to be inefficient, to use memory poorly, and the cry goes up for bigger, faster machines! When the machines are procured, even larger hunks of data are indiscriminately shoved through black box implementations of algorithms in hopes that meaning will emerge on the far side. It never does, but maybe with a bigger machine… I spent five years working in bioinformatics, and this is exactly the a…

Hi, I'm a bioinformatics researcher. Apparently I work for this guy's ex(?)-employer although I have never heard of him before.

My single most limited resource is programmer time. My time and the time of other people who work with me. I have access to loads of computers that sit idle all the time, even if it is on nights and weekends. There is zero opportunity cost to me in using these computers more fully. I have enough human work to do that I can wait for the results without having any wait states.

There can be a big opportunity cost in trying to rework a workflow so that it is more efficient and then test it thoroughly ensure correctness. Doing this may seem more appealing to someone who is interested primarily in computational efficiency. But I am more interested in research efficiency, and so are my employers and funders.

Re: A farewell to bioinformatics (2012)

#22

Post by hnwh » Sun, Jan 27, 2013, 5:08 AM UTC

michaelhoffman wrote: post #21

zerohp wrote: post #6

> the software is written to be inefficient, to use memory poorly, and the cry goes up for bigger, faster machines! When the machines are procured, even larger hunks of data are indiscriminately shoved through black box implementations of algorithms in hopes that meaning will emerge on the far side. It never does, but maybe with a bigger machine… I spent five years working in bioinformatics, and this is exactly the a…

Hi, I'm a bioinformatics researcher. Apparently I work for this guy's ex(?)-employer although I have never heard of him before. My single most limited resource is programmer time. My time and the time of other people who work with me. I have access to loads of computers that sit idle all the time, even if it is on nights and weekends. There is zero opportunity cost to me in using these computers more fully. I have en…

How can you leverage all of us really good programmers with tons of time, who are dying to work on something "important" and meaningful?

Re: A farewell to bioinformatics (2012)

#23

Post by michaelhoffman » Sun, Jan 27, 2013, 5:10 AM UTC

gwern wrote: post #18

jmspring wrote: post #8

Sounds like a fed up academic with a stick up his backside. Sh*tty data? Comes from the community. If the data and algorithms are so poor, and the author so superior, he should have been able to improve the circumstances. This whole screed reads like an entitled individual who entered a profession, didn't get the glory, oh and yeah, academia doesn't pay well. In the realm of bioinformatics, lets ignore the work done…

> Sh*tty data? Comes from the community. If the data and algorithms are so poor, and the author so superior, he should have been able to improve the circumstances. Why? Aren't you assuming a lot about the incentives? What if the ground truth is simply that all the results are false due to a melange of bad practices? Do you think he'll get tenure for that? (That was a rhetorical question to which the answer is 'no'.)…

He's not getting tenure because he doesn't have a PhD. According to LinkedIn, he has a master's degree awarded after four years of study [1], which often indicates someone who did not complete a PhD.

[1] http://www.linkedin.com/pub/frederick-ross/13/81a/47

Re: A farewell to bioinformatics (2012)

#24

Post by jostmey » Sun, Jan 27, 2013, 5:10 AM UTC

Basic science moves forward slowly limited by the pace of fortuitous discoveries. I have found that many people from the field of computer programming have unrealistic expectations of what can be done in biology and other sciences.

Re: A farewell to bioinformatics (2012)

#25

Post by zerohp » Sun, Jan 27, 2013, 5:11 AM UTC

leoh wrote: post #10

zerohp wrote: post #6

> the software is written to be inefficient, to use memory poorly, and the cry goes up for bigger, faster machines! When the machines are procured, even larger hunks of data are indiscriminately shoved through black box implementations of algorithms in hopes that meaning will emerge on the far side. It never does, but maybe with a bigger machine… I spent five years working in bioinformatics, and this is exactly the a…

Part of the problem is grant money. Sometimes it's faster to buy more machines and get more results as opposed to rewriting entire algorithms. But the author does correctly identify, I think, some tendencies of some academic bioinformaticists.

I have enough experience to know if this is true or not. Many times it was faster to buy more machine, but often it was not. We already had 10000 cores.

I proposed, implemented, and tested an 8 line change to our alignment tool that saved 6% cpu time. It took me two days, most of which was my spare time at home. This one program was using 15 cpu years every month. Nobody cared. It never went into production. I started interviewing for a new job and left shortly after that.

Re: A farewell to bioinformatics (2012)

#26

Post by dinkumthinkum » Sun, Jan 27, 2013, 5:11 AM UTC

Why is this on the front page or why is it relevant? It's kind of a rant. I did some work on a publication in this field and was published once; I don't think it is a horrible research program. There may exist some of the issues in bioinformatics described here but I don't think it is terribly productive.

Re: A farewell to bioinformatics (2012)

#27

Post by kylemaxwell » Sun, Jan 27, 2013, 5:16 AM UTC

You know, I'd be more inclined to listen to him if he didn't also completely decry almost all of modern biology, which (in my view) has been to the late 20th and early 21st centuries what physics was to the late 19th and early to mid 20th centuries.

Re: A farewell to bioinformatics (2012)

#28

Post by michaelhoffman » Sun, Jan 27, 2013, 5:18 AM UTC

hnwh wrote: post #22

michaelhoffman wrote: post #21

Earlier quoted context omitted.

Hi, I'm a bioinformatics researcher. Apparently I work for this guy's ex(?)-employer although I have never heard of him before. My single most limited resource is programmer time. My time and the time of other people who work with me. I have access to loads of computers that sit idle all the time, even if it is on nights and weekends. There is zero opportunity cost to me in using these computers more fully. I have en…

How can you leverage all of us really good programmers with tons of time, who are dying to work on something "important" and meaningful?

If you want to improve the software engineering quality of bioinformatics software, then find an open source project you are interested in, and offer to submit patches to improve really unsexy but important stuff for bioinformatics user experience. Things like documentation, deployment, user interface, and testing. Some of these things require little domain knowledge but no one wants to do them.

Edited to add: some projects might even have a bug tracker that will already have problems you can tackle.

Re: A farewell to bioinformatics (2012)

#29

Post by MartinCron » Sun, Jan 27, 2013, 5:19 AM UTC

droithomme wrote: post #5

mscarborough wrote: post #2

>> I’m leaving bioinformatics to go work at a software company with more technically ept people and for a lot more money. More money, good on you. Starting off your critique of your former colleagues with "technically ept people'...not going to get a lot of sympathy for the correctness of your work.

Well, ept is obviously a back-formation and a clever and amusing one. Etymology is straight from Latin: ineptus, which is prefix in- plus aptus (fitting or suitable). Interestingly there's also inapt which is quite similar. edit: aheilbut's research on this is much more thorough.

I used a similar back figuring when describing a co-worker who was in the wrong job... "He's not inept, he is inapt"

Re: A farewell to bioinformatics (2012)

#30

Post by aheilbut » Sun, Jan 27, 2013, 5:24 AM UTC

I sympathize with the author, but this piece fails because many of the specific criticisms are off-base, and he's not trying to be at all constructive.

For example, it isn't true at all that microarray data is worthless. The early data was bad, and it was very over-hyped, but with a decade of optimization of the measurement technologies, better experimental designs, and better statistical methods, genome-wide expression analysis became a routine and ubiquitous tool.

The claim that sequencing isn't important is ridiculous. It's the scaffold to which all of biological research can be attached.

However:

There is a great deal of obfuscation, and reinventing well-known algorithms under different names (perhaps often inadvertently). There's also a lot of low-quality drivel on tool implementations or complete nonsense. This is driven largely by the need in academia to publish.

The other side of this problem is that in general, CS and computer scientists don't get much respect in biology. People care about Nature/Science/Cell papers, not about CS conference abstracts. Despite bioinformatics/computational biology not really being a new field anymore, the cultures are still very different.