Re: Question about Maker unmask optionSimple repeats are soft masked (lower case), so they are still visible to the gene predictor. It is the high complexity repeats that are hard masked (‘NNNNN’) as they can cause havoc with gene predictors (they ususally encode real proteins, retrotransposase etc.).
Soft masking the simple repeats has the benefic of not seeding alignments in those regions while allowing alignments to extent through them. This avoids spurious BLAST alignments, and it leaves the sequence available for gene prediction
So setting unmask to 1 basically makes retrotransposon regions visible to the gene predictor. Not really a good idea, but can sometimes be useful in situations where you believe that transoposons might have been integrated into the real gene structure (rare).