Which are the best approaches for searching oligonucleotides (6-25 bp) in DNA sequence collection (like human genes set) allowing mismatches, insertions and deletions?
for a sequence length of so tiny (6nt) you bound to have a huge number of non specific hit, and it would be difficult to say from where is it actually coming from? In general practice people use BWA algorithm for alignment of such short sequences. I think you can use bowtie/ BWA for your purpose.
It's the strategy that i'm already using but these algorithms are very useful only if you have a single target sequence or if you want to know all the positions of the pattern.
Instead, if you just want to know which sequence contains the pattern (at least once) they can be very inefficient.
I'm searching for something like a DNA database framework (there are many) but that work with pattern matching. Usually, database allow to find strings similar to a query string, but not substrings.