% sequence similarity