Shell script to search one file for contents of another and replace text?
 

Hi all,

Suppose I have a pair of files containing lists. The first file is called contigs.txt and it contains a list that looks like this:
The second file is called reads_in_contigs.txt. It consists of a row of *s, a Contig name, a second row of *s, and a list of the singletons that make up that contig. It looks like this:
What I want to do is search through the reads_in_contigs.txt file and replace contig names with the singletons that make them up. For example, I would like to replace the text "Contig822" in the reads_in_contigs.txt file with the following:
Any suggestions would be greatly appreciated!

Thanks,
Kevin

And when you use contigs.txt ?

Hi,

do you mean something like this
It is not clear from your initial post how you want to handle non-existent replacements. E.g., Contig826 has no replacement according to your sample data. Do you want to keep it or replaced it with nothing? The above script keeps it. The resulting file is named 'result'. Rename it if necessary.

I am with goosen in that the following does not mention the contigs.txt file anywhere:
Although I see crts' crystal ball is working better than mine ;)

crts, Nailed it! This really helped me out. Thank you. As you guessed, if a Contig# doesn't appear in reads_in_contigs.txt (e.g., Contig826) I want it to remain unchanged in contigs.txt. All of the Contig# names in contigs.txt should have been replaced with the singletons that make them up so if one remained it would serve as a red flag to me that I did something wrong.

grail and goosen, I'm sorry I wasn't very clear on what I was trying to do. If you care, it should have said: "What I want to do is search through the reads_in_contigs.txt file for matches to contig and singleton names in the contigs.txt file and replace contig names (e.g., Contig822) with the singletons that make them up." Does that make more sense?

Thanks a lot!
Kevin

Cool ... in that case how about something like:

Hi all,

I'm trying to do something similar to the above but with a twist. Say we have a similar contigs.txt file that looks something like this:
We also have a file containing DNA or amino acid sequences (sequences.txt) where each sequence spans two lines. The first line that begins with ">" is the header (some of these will correspond to entries in contigs.txt) and the second is the sequence itself.
I want to extract the headers and sequences in the sequences.txt file and send them to a new file if the header (not including the ">") appears in the contigs.txt file. I thought an awk command might work well but I can't figure out how to use getline to get multiple lines. Also, it won't print the ">" that I want to add in. Any help would be greatly appreciated! Here's what I have so far: