Optimizing micropattern geometries for cell shape and migration with genetic algorithms

Adhesive micropatterns have become a standard tool to control cell shape and function in cell culture. However, the variety of possible patterns is infinitely large and experiments often restrict themselves to established designs. Here we suggest a systematic method to establish novel micropatterns for desired functions using genetic algorithms. The evolutionary fitness of a certain pattern is computed using a cellular Potts model that describes cell behavior on micropattern. We first predict optimal patterns for a desired cell shape. We then optimize ratchet geometries to bias cell migration in a certain direction and find that asymmetric triangles are superior over the symmetric ones often used in experiments. Finally we design geometries which reverse the migration direction of cells when cell density increases due to cell division.