在文件中划分两列并将新列中的输出打印到多个文件的同一文件中

Question

I have a number of files which is in VCF format.That is how it looks like

1   127573  rs7 G   A   79.78   .   AC=1;AF=0.500;AN=2;BaseQRankSum=1.231;ClippingRankSum=-0.358;DB;DP=5;FS=3.979;MLEAC=1;MLEAF=0.500;MQ=60.00;MQ0=0;MQRankSum=0.358;QD=15.96;ReadPosRankSum=1.231  GT:AD:DP:GQ:PL  0/1:2,3:5:27:108,0,27

In which i need to divide the second part of last column and and print the output in new column.. ie, from the above example, its 3 and 5 ( from 10th column 0/1:2,3:5:27:108,0,27) and the output it should look like, That is with 0.6 (ie 3/5) as last column

 1  127573  rs7 G   A   79.78   .   AC=1;AF=0.500;AN=2;BaseQRankSum=1.231;ClippingRankSum=-0.358;DB;DP=5;FS=3.979;MLEAC=1;MLEAF=0.500;MQ=60.00;MQ0=0;MQRankSum=0.358;QD=15.96;ReadPosRankSum=1.231  GT:AD:DP:GQ:PL  0/1:2,3:5:27:108,0,27 0.6

In order to achieve this I used awk in unix, as follows,

cat result_1 |cut -f10 | sed 's/:/\t/g' >sample
cat sample | cut -f2 | sed 's/,/\t/g' | awk '$2!=0 || $3!=0{print $1"\t"$2"\t"$2/$3}' >result_1 

But it complains as

awk: (FILENAME=- FNR=1) fatal: division by zero attempted

any other alternative solutions in Python or Perl would be great..!!!

Answer 1

awk '{split($NF, a, /[,:]/); $(++NF) = a[3]/a[4]; print}' file

好的，除以零：

awk '{split($NF, a, /[,:]/); $(++NF) = (a[4]==0 ? "Inf" : a[3]/a[4]); print}' file

Answer 2

Here's one perl way of doing it:

perl -ne 'chomp;if(/\t[^, ]+,(\d+):0*([1-9]\d*)[\S ]*$/){$n=$1;$d=$2;print("$_\t",$n/$d,"\n")}else{print("$_\t\n")}' < result_1 > result_1.new

This will do it. It will ensure a non-0 positive value for the denominator in the match ([1-9]\\d*), and allows for leading zeros with the '0*' in front of it.

The chomp removes the hard return ("\\n"), so it's tacked on in the print.

It ensures you're parsing the last column from the last tab to the end of the string and it allows spaces.

The -n wraps the code in while(){...}.

It adds a tab even if there would have been a division by zero but in that case, leaves the last column empty.

You can mv the file afterward if you want to overwrite the original, but I prefer to save precursors as a backup.

There probably exists a more succinct/readable way of doing it in perl or via another language, but this suffices.