Dear all, I am trying to write a simple script that takes as input a list of identifiers, searches in a large text file for each of those identifiers, and where it finds them, replaces their occurence in the text file as ID_1, ID_2, ID_3, etc. The entries in the large text file look like this: >probe:HG_U95Av2:1138_at:395:301; Interrogation_Position=2631; Antisense; TGGCTCCTGCTGAGGTCCCCTTTCC >probe:HG_U95Av2:1138_at:322:441; Interrogation_Position=2661; Antisense; GGCTGTGAATTCCTGTACATATTTC >probe:HG_U95Av2:1138_at:213:419; Interrogation_Position=2703; Antisense; GCTTCAATTCCATTATGTTTTAATG where "1138_at" is the identifier. I would like the output file to be identical to the input file, except that 1138_at would be replaced by 1138_at_1, 1138_at_2, 1138_at_3, etc. The code below finds and replaces all instances of 1138_at with 1138_at_1; that is, the counter does not increment. Does anyone have any suggestions? Thanks very much. ~ MikeO CODE: exec<"All_IDs.txt" i="1" while read line; do echo $line sed -e "s/:$line/&_$i:/" Large_file.txt >>Output.txt i=$[$i+1] done

There is no simple way of performing this substitution. Basically, I count the number of substitutions that must be made. Then, use sed inside of a loop to perform one substituition at a time: #!/bin/ksh typeset -Z2 x cp data.file tmp.data.file cnt=$(grep -c 1138_at tmp.data.file) x=0 while [ $x -lt $cnt ] do x=$((x+1)) sed -e '1s/1138_at/1138_'"$x"'/;t' -e '1,/1138_at/s//1138_'"$x"'/' tmp.data.file > /tmp/mytmp.file mv /tmp/mytmp.file tmp.data.file done The final solution will be in tmp.data.file.