> IMHO there is a lot of low-hanging fruit for startups in the field of symbolic AI applied to biology and medicine.

I think the issue in this area is mostly to convince and sell to bureaucratic institutions.