We might not know enough, but there's much we CAN do even with today's limited knowledge. I see this field moving so fast that in relative few years we could do almost anything in living things.
Omri (CEO Genome Compiler)
Needs simulation stages for testing and debugging :) I can just imagine the abominations created by an off-by-one error or the Lovecraftian horrors made possible through a SQL injection vulnerability.
Although it looks cool, it seems to me that our current theoretical background in biology makes this 'genome compiler' irrelevant. We simply don't know enough.