UNIX for Dummies Questions & Answers

Hello,
I have a file with 5 columns that looks like this:

A1BG chr19 + 58863335 58866549
A1BG chr19 - 58858171 58864865
A2LD1 chr13 - 101182417 101186056
A2LD1 chr13 - 101182417 101241046
A2M chr12 - 9220303 9268558
A2ML1 chr12 + 8975149 9029377

I want to subtract the value in column 5 from column 4 and send the results to a 6th column and the file to look like this (6th col = values in 5th- values in 4th

A1BG chr19 + 58863335 58866549 3214
A1BG chr19 - 58858171 58864865 6694
A2LD1 chr13 - 101182417 101186056 3639
A2LD1 chr13 - 101182417 101241046 58629
A2M chr12 - 9220303 9268558 48255
A2ML1 chr12 + 8975149 9029377 54228


Then I want the file to get rid of the duplicates (I already used the sort and uniq commands and I still have some left e.g A1BG, A2LD1)
I want my final file to look like this, with only the genes with greatest length recorded (as below: only the largest values for genes A1BG and ALD1 were recorded:

A1BG chr19 - 58858171 58864865 6694
A2LD1 chr13 - 101182417 101241046 58629
A2M chr12 - 9220303 9268558 48255
A2ML1 chr12 + 8975149 9029377 54228

Please help!
Thanks so much!

Hi,

Try this one,

Code:

awk '{a=$5-$4;print $0,a;}' Input_File

For unique lines try this one,

Code:

awk 'BEGIN{re=0;}{if(re != $1 ){a=$5-$4;print $0,a;re=$1;}}' Input_File

Cheers,
Ranga

The following script only prints the instance of any duplicate with the highest value of the new final field.

Code:

perl -e 'while(<>){
   chomp;
   @fields=split(/ /,$_);
   if (($fields[4] - $fields[3]) > $genes{$fields[0]}->[5]){
      $genes{$fields[0]}=[@fields[0..4],$fields[4] - $fields[3]];
   }
}
for (sort keys %genes){
   print join (" ", @{$genes{$_}}),"\n";
}' tmp.dat

To fulfill wolf request, modified ranga's code

Code:

$ nawk '{a=$5-$4;print $0,a}' infile | sort +5nr | nawk '!x[$1]++'
A2LD1 chr13 - 101182417 101241046 58629
A2ML1 chr12 + 8975149 9029377 54228
A2M chr12 - 9220303 9268558 48255
A1BG chr19 - 58858171 58864865 6694

Sorry to be confused on this one but where do I enter my file name? at the end of:
nawk '{a=$5-$4;print $0,a}' infile | sort +5nr | nawk '!x[$1]++' myfile.txt Am I doing this correctly?
Sorry, I'm really new to Unix.

Code:

nawk '{a=$5-$4;print $0,a}' infile | sort +5nr | nawk '!x[$1]++'

Your file is specified in red.

Then the pipe operator "|" will branch the output of the previous command as input of next command.

so the output of

Code:

nawk '{a=$5-$4;print $0,a}' infile

will be given as input of

Code:

sort +5nr

then output of

Code:

nawk '{a=$5-$4;print $0,a}' infile | sort +5nr

will be given as input of

Code:

nawk '!x[$1]++'