Accelerated reference alignment (via something like an FPGA) is definitely a way forward as the problem is now relatively well understood. Would be interesting

Accelerated reference alignment (via something like an FPGA) is definitely a way forward as the problem is now relatively well understood. Would be interesting to see where things go with respect to de-novo and graph alignment.