The main idea of this project was to classify proteins by
homotopy, which would be inbetween the geometry (the exact shape
of the protein) and the topology. The project was put aside after
a while.
Under ideal conditions, there are algebraic relationships in
multi-locus linkage disequilibria that I thought might be
exploited. Gradually, out of necessity, the attention shifted more
towards how population structures could be identified, modelled or
corrected for.