Thank you for your answer.
My only concern is the following:
Let us say I have a genome
ACAGTACA
and I snip away the GTA part before building an index. Won't the index now tell me that the last nucleotides in the original string begin at the wrong position?
Ie original string and the corresponding index is
ACAGTACA 12345678
When I snip away GTA and index that I get
ACACA 12345
but of course, what I want is
ACACA 12378
How many short sequences do you have? Do you need to capture mismatches, or are exact matches only in the genome acceptable?