I have used Tandem Repeats Finder (TRF) for tandem repeat search in my fasta files.
Output looks like this:

Sequence: ENSG01

Parameters: 2 5 7 80 10 50 2000

1053 1139 4 22.2 4 67 2 62 28 4 48 18 1.68 GAGT GAGAGAGTGGGTTAGAGAGTGAGTGAGCCAGTGAATGAGTGAGTG
1069 1137 20 3.6 19 74 7 71 28 5 47 17 1.70 GAGTGAGTGAGCCAGGAGT GAGTGAGTGAGCCAGTGAATGAGTGAGTG
1619 1746 8 16.8 8 65 9 60 27 1 52 18 1.55 GAGT GAGTGAGTGAGTGAATGAGTGAATGGGAGT

Sequence: ENSG02

Parameters: 2 5 7 80 10 50 2000

Example explanation:

Sequence (ENSG01) - fasta name
Column 14 (GAGT) - repeat unit
Column 15 (GAGAGAGTGGG) - repeat sequence

Help I need:

• How to process such file:

  • Remove sequences that don't have repeats (like ENSG02) in them?

  • Combine fasta name with the following repeat data?

    For the output like this:

    ENSG01
    1053 1139 4 22.2 4 67 2 62 28 4 48 18 1.68 GAGT GAGAGAGTGGGTTAGAGAGTGAGTGAGCCAGTGAATGAGTGAGTG
    1069 1137 20 3.6 19 74 7 71 28 5 47 17 1.70 GAGTGAGTGAGCCAGGAGT GAGTGAGTGAGCCAGTGAATGAGTGAGTG
    1619 1746 8 16.8 8 65 9 60 27 1 52 18 1.55 GAGT GAGTGAGTGAGTGAATGAGTGAATGGGAGT
    

    I guess it's possible to grep '^[0-9]', but I don't know how to join such grep output with the fasta name.

• My hypothesis is that tandem repeats has DNA motif like structure. How can I use repeat unit (like GAGT) and search for such motif occurrences genome wide?

  At the moment my plan is:

  • Every unit GAGT has it's sequence of occurrences GAGAGAGTGGGTTAGAGAGTGAGTGAGCCAGTGAATGAGTGAGTG
  • Submit GAGAGAGTGGGTTAGAGAGTGAGTGAGCCAGTGAATGAGTGAGTG to MEME and get PSPM

  • Scan PSPM genome wide

    I am using MEME as I don't know how can I scan for a given unit GAGT allowing mismatches.

    What should I do with redundant repeat units? For example: GAGT and GAGTGAGTGAGCCAGGAGT. Should I use the shortest unit like GAGT or all units even they overlap?

Thank you for you time.