Chip-chip data analysis - oh I wish I were a bioinformatician! (Jun/17/2008 )

Pages: Previous 1 2

QUOTE (toejam @ Jun 24 2008, 06:42 PM)

oh well...60 is a lot more than i expected...

a script to compare them all should do.

yes, it finds differences between the files

that's not what you want....

Yup - I want similarities... oh well..
Thanks anyway!

-Clare-

hello again clare,

i'll be completly honest with you. i have no idea how to do that on windows. to do those kind of analysis i usually use the linux partition in the computer. i installed a distro called "ubuntu", it's one of the most user friendly there are. in the terminal of linux the command "sort" as i pointed out before, can do what you are looking for and i believe it could do it with the 60 + experiments you'd like to analyze, sorry "analise" for the british english

if there's a server at your university/institute running linux and you can get an account, you could use the ssh-server application (also available for windows) to enter the terminal and run the program there. otherwise, i suppose with the activeperl you could do it but that is beyond my knowledge. best luck.

-toejam-

QUOTE (toejam @ Jun 24 2008, 11:30 PM)

if there's a server at your university/institute running linux and you can get an account, you could use the ssh-server application (also available for windows) to enter the terminal and run the program there. otherwise, i suppose with the activeperl you could do it but that is beyond my knowledge. best luck.

Eeep! Ok then

I think it's time to call in the experts! I'll have a chat to the guys who wrote our R script and see what they suggest

-Clare-

hey clare,

have you consulted the experts yet? i just remembered (by the new look of bioforum) that mac's system is based on unix, that means you can use a program called "terminal" (under utilities) where you can do stuff like the one mentioned here. i think it's easier to get a mac than another operating system.

-toejam-

Hi All,

TJ, I don't think your command will work on files like that (unless you want to order by chromosome).
If thinks really are that simple ! you could do something like this (untested)

If you combine all your experiments into one file. Each experiment with a distinct number in front:
1 Chr12 1000-1500
1 Chr5 600-700
...
...
...

CODE

#!/usr/bin/perl  #yuck I hate perl
use warnings;
use strict;

open F, q/my_file_name_goes_here/ or die $!;
my %HASH = ();

while(<F>) {
   chomp;
   my ( $exp, $data ) = split /\s+/, 2;
   push @{ $HASH{$data} }, $_;
}

for ( keys %HASH ) {
  print "@{ $HASH{$_} }\n";
}

You would end up with similar results with their experiment identifier.

-perlmunky-

Cheers for the code but I think my analysis will be a little more complicated! I have had a closer look at my results and the genomic regions aren't always the same. ie: in one patient it might be Chr1: 1000-2000 and in another Chr1: 850-1700.
I have a meeting with the bioinformaticians next week so fingers crossed! Will post if we figure it out

-Clare-

Pages: Previous 1 2