Need help with a file that prints letters from a file according to another file!

kylle345 · #1 (**permalink**) 05-09-2009

So basically what I want to do is pull out DNA sequences for a particular gene name.

I have 2 files (FILE1 and FILE2) and I want an output into a separate file (FILE3).

FILE1 and 2 are MASSIVE so I am only posting examples from each file.

So FILE1 looks like this (tab deliminted, 4 columns):

##gff-version 1

1154 10 + AAD6
418 7429 + AAH1
702 759 + AAT1
584 10 - ABF2
642 4894 - ACC1
651 7213 - ACN9
1055 3454 - ADE1

The next file, FILE2, looks like this:

>1154
ATCTCACTCGTAATTCTACATAATTTTGTTTATGCTTTTATTGTCATTTTATATATTGTCAGTCATTATCCTATTACATTATCAATCCTTGCATTTCAGC TTCCACTTATTTCGATGACCGCTTCTCATAACTTATGTCATCTTCTAACACCGTATATGATAATGTACCAGTAGTATGAC
>584
GCAAGCTTTATAGTGACAACAATAAGGTATCACTCGGTTACAATTACCCCCACTTCCCCT

What I want to do is identify column 1 of FILE1 with the ># on FILE2. So for example, 1154 from FILE1 will match up with 1154 from FILE2. Next, I want it to identify the value on column 2 (so for 1154, it will identify the 10th letter which happens to be G). So if column 3 of FILE1 is + then it will print the first 8 letters in from of it (i.e. the 8 letters in front of G would be TCTCACTC). But if is it – on column 3, then it will take the reverse. So for ABF2 on “584” it will take the top 8 sequences starting from the reverse end. So instead of starting at “G” at >584, it will start at “T” (the end). So the position of ABF2 will be 25 letters away from “T” , so the letter will be “C”. Then it will take the values behind it… so CCACTTCC.

The output file will print out column 4 of FILE1, the top 8 letters from FILE2 and column 3 from FILE1.

The final file (FILE3) will look like this:

AAD6 TCTCACTC +
ABF2 CCACTTCC -

Could someone give me some help on this! I am new to perl and I am put in a situation where I have to program at a very high level.

Thanks

devtakh · #2 (**permalink**) 05-09-2009

I am not clear on this part -

HTML Code:

But if is it – on column 3, then it will take the reverse. So for ABF2 on “584” it will take the top 8 sequences starting from the reverse end. So instead of starting at “G” at >584, it will start at “T” (the end). So the position of ABF2 will be 25 letters away from “T” , so the letter will be “C”. Then it will take the values behind it… so CCACTTCC.

devtakh · #3 (**permalink**) 05-10-2009

Try this.

Code:

awk 'FNR==NR{a[$1]=$2SUBSEP$3","$4;next}
/^>/{gsub(/>/,"",$1);s=$1;
if (s in a){
getline;
st=substr(a[s],1,index(a[s],SUBSEP)-1)
sg=substr(a[s],index(a[s],SUBSEP)+1,1)
if ( sg == "+")str=substr($0,st-8,8);else str=substr($0,length($0)-st,8);pt=substr(a[s],index(a[s],",")+1,length(a[s]))
print pt,str,sg;next
}}' file1 file2

cheers,
Devaraj Takhellambam

ghostdog74 · #4 (**permalink**) 05-10-2009

how MASSIVE is your file1 and file2, in terms of MB?? GB??
also, if you are new to Perl, then you should at least read up something on Perl before attempting this.

More UNIX and Linux Forum Topics You Might Find Helpful
Thread	Thread Starter	Forum	Replies	Last Post
file size comparision local file and remote file	dba.admin2008	Shell Programming and Scripting	4	11-13-2008 05:57 PM
Reading a file and writing the file name to a param file.	thebeginer	UNIX for Advanced & Expert Users	1	10-05-2007 04:38 PM
Reading file names from a file and executing the relative file from shell script	anushilrai	Shell Programming and Scripting	4	03-10-2006 05:25 AM
How can I find the 3 first letters from the name file	steiner	Shell Programming and Scripting	8	06-17-2005 08:10 AM
look in file, seperate letters, put in order...	chekeitout	UNIX for Advanced & Expert Users	3	11-05-2004 05:00 PM

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes	Rate This Thread
Switch to Linear Mode Hybrid Mode Switch to Threaded Mode	Rate This Thread:


	Powered by