Script to bring together all relevant accession into its root access

Question

I have a big data following;

KEL_1021159,K00001
KEL_1020176,K00001
KEL_1018609,K00001
KEL_1008140,K00006
KEL_1012058,K00006
KEL_1018645,K00006
KEL_1004034,K00006
KEL_1004235,K00006

and I am currently trying to convert it to like this;

KEL_1021159,KEL_1020176,KEL_1018609 K00001
KEL_1008140,KEL_1012058,KEL_1018645,KEL_1004034,KEL_1004235 K00006

Is there any basic script for such a purpose?

Answer 1

Pretty straightforward with awk:

awk -F, '
    {a[$2] = a[$2] $1 FS}
    END {
        for (key in a) {
            sub(/,$/, "", a[key])
            print a[key], key
        }
    }
' file

Opaque perl one-liner:

perl -F, -ane 'push @{$a{$F[1]}},$F[0]}{print join(",",@{$a{$_}})." ".$_ for keys %a' file