Wednesday, January 23, 2008


Recovering the cursor

Hi everyone,

this tip is for someone who, like me, usually shuts down the graphical interface of some openGL programs like O or Dino without issuing a 'quit' command.

All you have to do is executing 'reset' at the command line prompt, after exiting your O session. You will have to type it blindly since you don't have any visual feedback.

Tuesday, May 29, 2007


How to add swap space to a running Linux kernel

I've found this to be quite interesting when working with a virus particle, even for the smallest of the picornaviruses. You just have to perform some commands, and your virtual memory gets multiplied to tackle those huge structural problems.

First of all, you need to be root user for a brief time.

cd /var/tmp <- Go to the area where the temporary swap space will be located.
dd if=/dev/zero of=myswap1 bs=1024 count=1048576 <- Create a '0'-filled archive.
mkswap -c -v1 /var/tmp/myswap1 <- Actually make the swap space.
swapon /var/tmp/myswap1 <- Activate the swap space and add it to the kernel memory mapping facility.

You can make as many swap files as you need. I don't know if their size has a limit.

Wednesday, October 04, 2006


How to retrieve the aminoacidic sequence (Fasta format) from a PDB

This trick is an easy one, and is useful. Just make sure that you have 'pdbset' correctly installed and accesible in $PATH.

pdbset xyzin pdbfile << EOF

Good luck!

Friday, September 29, 2006


Ultimate weapon of crystal annealing

Following Terese Bergfors' advice of using a playing card for performing a subtle crystal annealing, I've managed to get this "pretty" picture:

Result: No enhanced spots. No better resolution. It didn't even improve my Poker skills.

EDIT: After trying it several times, and exchanging cards with my team mate, I've finally *consistently* improved my resolution, from the initial 28 Angstrom to a not-so-poor-but-still-pretty-useless 18 Angstrom. Woohoo. Let's see if a genuine synchrotron ride (to hopefully happen next week) makes these protein conglomerates really deserve the name Krystallos.

And, yes, this method is too cool for the school.

Friday, June 23, 2006


Molecular surfaces

Hi there,

I've found an interesting page about molecular surfaces calculations. Please be sure to check it out if you're interested in that topic:

Click me! You won't be disappointed

Thursday, March 30, 2006


High resolution pictures in Pymol 0.99

Hi there, true believers!

Warren DeLano has shared with [ccp4bb] a simple way for creating publication apt images right from the Pymol 0.99 command line, skipping the use of image editing programs such as Gimp, Krita and so. I've found it to be very interesting and entertaining, so I'll reproduce it here:

# for OpenGL images:

draw 1200,900
png image1.png, dpi=300

# for ray-traced images:

ray 1200,900
png image2.png, dpi=300

Hope it's useful for you!

Friday, March 24, 2006


Introducing pdbsplit 0.3.1

Hi comrades,

I've finished version 0.3.1 of my handy utility called pdbsplit, which can right now split a single PDB file into several chains, domains, aminoacid groups, etc. You can now choose different ranges of aminoacids to be output, and there switches for ripping headers, ligands, waters, etc.

The download link (Unix sources) is here.


This page is powered by Blogger. Isn't yours?