So if I understood you correctly then further lowering the cost of experimentally determining protein targets could be a viable way forward that is completely orthogonal to computational methods?