I have blogged a bit about tools for parallelising R code in the past, and back in October I was interviewed for Genomeweb about my thoughts on parallel computing and R/parallel.
I completely missed the piece when it came out, though, but I stumpled upon it today: Making the most of multicore. (login required)
Go read it!
--
52-73=-21
Yesterday, I mentioned that I was going to two seminars at BiRC on Darwinism and ID...
RBH Says:
February 20th, 2009 at 1:13 am I’m looking forward to your comments on the seminar on intelligent design (which I am coming to think of as “intentional design,” not so intelligent).
Ok, here goes :)
The first talk was by a historian and was on the history of Darwin and the introduction of his ideas in Denmark. Here it was pretty rapidly accepted by scientists (and the population in general). One of the reasons for this was that Danish theologians quickly made a distinction between "knowing" and "believing" and had no problems with natural sciences dealing with science and the church dealing with faith, and not trying to mixing that up.
This is still the case today, but at the end of the talk he turned to the history of creationism and ID. These movements are still pretty rare in Denmark, so it was mainly the history of creationism in the US.
All in all a very interesting talk.
The second talk was from a philosopher and was on the arguments used by ID, and whether ID should be considered a science and taught as such in school.
The arguments are well known to you, I am sure. Argument by analogy, god of the gaps, and the two chestnuts Irreducible Complex and Specified Complexity. (The two latter he essentially reduced to God of the Gaps by showing that the final step for Design in either case boils down to not knowing the "true" answer and therefore conclude that "God did it"; regardless of the many complexity/probability theoretic problems with the arguments, that alone disqualifies it as proper logic).
Arguing by analogy is essentially the "if it looks like a duck, it must be a duck" argument, and the problem is that by choosing the right analogy you can prove anything. He quoted the Danish play Erasmus Montanus where the main character proves that his mother is a stone, since stones cannot fly and neither can his mother.
The God of the Gaps argument essentially asks you to conclude that if you don't know how X happens, then it must be God who did X. If you cannot prove not X, then X must be true. The problem here is, of course, that you can just plug in Y for X and the argument is just as (in)valid. So maybe the Spaghetti Monster did it?
All in all, he concluded that non of the arguments would be considered strong arguments (from a philosophy view point).
He then turned to ID as a science, and compared it to the scientific principle of falsifiability. If God did it, and he works in mysterious ways, then falsifiability is out the window.
Sure, God might be behind it all, but then science is meaningless. He can cheat us at any time. You simply cannot have a mysterious God and science at the same time. If you accept science, then either God does not excist, or he does not work in quite as mysterious ways.
If you accept ID as science, then all other science has to go, since they are not philosophical compatible.
So should ID be taught in science classes? No!
--
51-72=-21
With the Neanderthal genome at 1.5x coverage, the Mammoth sequence in draft last year, and the Tasmanian tiger this january, I'm wondering how many extinct species have we sequenced yet?
Does anyone know?
What are the limits for extracting ancient DNA? Yeah, I know it depends on how well preserved the DNA is, but does anyone have rough numbers?
I wish I knew more about ancient DNA...
--
50-71=-21
How can you say that evolution is “true”? Isn’t that just your opinion, of no more value than anybody else’s? Isn’t every view entitled to equal “respect”? Maybe so where the issue is one of, say, musical taste or political judgement. But when it is a matter of scientific fact? Unfortunately, scientists do receive such relativistic protests when they dare to claim that something is factually true in the real world.
...
Scientific “truth” is only one kind (“Western” truth, the anthropologist may call it, or even “patriarchal”). Like tribal truths, yours merely hang together with the world view that you happen to hold, which you call scientific. An extreme version of this viewpoint (I have actually encountered this) goes so far as to say that logic and evidence themselves are nothing more than instruments of masculine oppression over the “intuitive mind”.
Grr... Does not compute!
Scientific “truth” is only one kind (“Western” truth, the anthropologist may call it, or even “patriarchal”). Like tribal truths, yours merely hang together with the world view that you happen to hold, which you call scientific. An extreme version of this viewpoint (I have actually encountered this) goes so far as to say that logic and evidence themselves are nothing more than instruments of masculine oppression over the “intuitive mind”.
Oh well, no more on this for now. This afternoon I'm attending two seminars: one on intelligent design and one on Darwin (both at BiRC, but with speakers from Dept. of Science Studies).
I might have more to say after that.
--
50-70=-20
Russ Altman has an interesting post on his blog:Bioinformatics & Computational Biology = same? No.
I spent the first 15 years of my professional life unwilling to recognize a difference between bioinformatics and computational biology. It was not because I didn’t think that there was or could be a difference, but because I thought the difference was not significant. I have changed my position on this. I now believe that they are quite different and worth distinguishing. For me,
- Computational biology = the study of biology using computational techniques. The goal is to learn new biology, knowledge about living sytems. It is about science.
- Bioinformatics = the creation of tools (algorithms, databases) that solve problems. The goal is to build useful tools that work on biological data. It is about engineering.
Personally, I have made the same distinguishing, but for some reason with the terms somewhat reversed. For me, computational biology has always been about the development of methods and tools, while bioinformatics has been about appyling methods to study biology.
I suppose someone can argue with the my use of the term “bioinformatics” as an engineering discipline. That’s fine–I’m open to a different term. But I would ask why bioinformatics isn’t good. I think computational biology is more solid–the ‘biology’ is clearly the noun and the ‘computational’ is clearly the adjective.
Good point for that use of the terms. My reasoning for the other use was that computational biology clearly had a focus on the "computational" and isn't just studying biology by running computer programs.
Anyway, the actual terms are not so important, but I completely agree that the mix of mathematics/statistics, computer science and biology -- whatever we call that mix -- consists of several disciplines:
I wouldn't put the first item, tools and methods development, entirely in an "engineering" box, though. Some tools development is just an engineering exercise, implementing existing well known methods, but some method development involves formalising new hypothesis and implementing ways of checking them into computer tools.
The same goes for the second point; applying tools can range from running data through a pipeline with very little other user involvement to detailed and careful analysis of all computational results compared to the underlying biological hypothesis.
--
49-69=-20