I think you can just to a replacement. In perl it would look like this

$seq =~ tr/ACGTacgt/TGCAtgca/;

From your other comment it looks like you're trying to look for self matching regions within a single strand, right?

A simple way I'd recommend is breaking down your sequence into fragments which you want to probe with, then searching against your sequence. Something like this:

from Bio import SeqIO
import re

for seq_record in SeqIO.parse(yourInputFile.fa, "fasta"):
    i = 0
    targetDict = {}
    while i < len(seq_record):
        targetDict[i] = seq_record.seq[i:(i+20)]
        i += 1

Then to search against your strand something like:

for pos in sorted(targetDict.keys()):
    t = targetDict[pos]
    if re.match(str(t), seq_record):
        print str(t) + "matches at " + pos

Just some parts pulled out of a script I use for something else, but hopefully will help as a starting point!