Archive for November, 2012

Funny question !

November 26, 2012 Leave a comment

Categories: Funny science

Funny prizes

November 7, 2012 Leave a comment

I ‘ve been pretty occupied recently, but now that things are better here are two awards that none shall miss.

The ig nobel price 2012:

As mentioned on the website the price award “achievements that first make people LAUGH then make them THINK”. My favourite are :

  1. The FLUID DYNAMICS PRIZE. There is no way someone could convince me that this kind of work is not absolutely useful. Every day while going back from the cafeteria in Agribio up to the office I face the puzzling question addressed by these authors and apparently I am not the only one hoping for eventual extension to other cups form
  2. The PEACE PRIZE. If ammunitions had any interesting use, I guess i could be this one
  3. The LITERATURE PRIZE. I wish the title doesn’t sound so familiar.


I sometime feels like a having an obsessive compulsive disorder regarding the presentation of codes (most often when the code was written by someone else and of course contains bugs that I may/should find)…Yes, I like commented code, self explanatory variable name, a nice indentation…so for sure when you have to compile this, you may fear the worst. The improbable but bare truth is that the code can be compiled (I don’t say that the compiler is not a bit worry) and last but not the least the executable works ! The participants really know what they are doing, I guess some of them should be professional code golfer

Software update

November 3, 2012 Leave a comment

Some updates :

Picard-Tools 1.79

Still an impressive update rythm (4 in less than 2 months), every time changes remains minor. (No worry)

GATK 2.2

This tool is reaching stability. In fact I used it to call variant on my project (Both unified genotyper and HaplotypeCaller). Despite some periods where updates came twice a day (a bit worry-some in fact), I had no real problems, fixes were mostly dealing with particular features or kind of data. So I guess we can consider this new version as mature enough.
Computations time is still a concern (You should not expect to get results on a big project within hours ;-), I hope that at some point multi-threading could work for all the tools (This would allow a better use of clusters).

ifort 13

To me the best fortran compiler, came up with few improvements, new Fortran 2003 et 2008 features, an enhanced MKL, and a better support for recent processors.

Cuda 5.0

The release promise efficiency improvements and a bunch of new features to ease development using CUDA. As a Fortran programmer, i am really eager to see alternatives to PGI compilers to easily take advantage of CUDA.


This project announced (and discussed) some time ago here and here and has been officially released at the end of august. This effort to come with both a new “method”¬† and accompanying tools should really be acknowledged. I tried stan while it was still in beta release and had no problem running¬† the examples. Despite compilation doesn’t frightened me at all, I believe the Rstan package is a good initiative as it may attract a wider audience. Some people had already played a bit here and there, with some classical example. I wish I could find time to do the same with an example in our field !

Categories: emacs, Linux, NGS