Title: Biological thought experiments and thecomputations they spawn
Abstract: Iwill talk about aspirations that bridge biology and computation. The first is determiningthe landscape of genetic contribution to autism. The second relates uncoveringpatterns of hominid mixing from present day sequence information. The third is speculations on evolution ofbiological systems.
(1) The Simons Foundation (SF) hascreated a rich resource for studying the genetics of autism, adevelopmental-neurological disorder, in the form of quad family whole genomesequence data. Although not easy, from enough of such data one could inprinciple determine the frequency with which new mutation and transmittedvariants contribute to autism. I will sketch out the genetic landscape in lowresolution now, with a discussion of how sharp it can become in the future. In particular, there are remaining challenges,paradoxes and enigmas. The path forward is largely (but not exclusively)computational and I will describe the kind of progress we can expect, and thekind of additional data needed.
(2) Quad family sequence data alsoprovide a possible portal into the human past, in the form of haplo-sequences.While not yielding a precise record of past genetic events, we may be able tosee into the past. For example, we might be able to determine if rates ofmutation have changed over recent human history, and we might see evidence oflarge scale intermixing of formerly isolated populations.
(3) Ant behavior can be modeled withfairly simple “algorithms”. We are beginning to investigate if we can implementthese as spatially-local algorithms that have rapid compute times andscalability. Thinking about the simplest ways to do this leads to speculationon genetic decay, evolution, tribalism, wars, and the molecular origins of self-recognition.I suspect we may discover the external conditions that drive species to undergoa phase transition to eusocial organization.