When looking for repeat patterns in a DNA sequence, we can look for the pattern and its reverse compliment with up to d mismatches. However, why do we look for reverse compliments if we're analyzing a single DNA strand?
I thought about it and my guess is that this lets us analyze both strands of DNA. For example, if we find the reverse compliment at position x, that means that the other strand of DNA has the pattern at position x - pattern.length.
Is that correct?
When we sequence genomes, the individual reads come from both strands, and that also applies for assembled sequences* (contigs / scaffolds). Therefore for most of the genome assemblies out there, strandeness of individual scaffolds is actually an arbitrary property, and reverse complementing any sequence will generate an equally valid representation (with reverted coordinates).
The coordinates of the reverse complementary k-mer matches depend on the notation. Usually, the notations are still in the coordinates of the reference, for example, blast gives for reverse complimentary strand matches the from subject coordinate greater than the to coordinate. Some others are denoting "the leftmost" position and strand encoded in a different column (e.g. sam).
*perhaps with the exception of chromosomal level assemblies
