please try to get your nomenclature at least halfway right

I agree with Michael (+1 above and below), but genes do have domains. Generally, gene domains, such as promoter, exon, intron, coding/non-coding, splice sites, etc. are either too general to require much of an algorithm or too esoteric to offer a complete and/or reliable prediction (eg, an entire promoter region). Add detail to your question to get a detailed, helpful answer.

I would guess (and I shouldn't have to) that unknown meant protein domains.