On the other hand, representing uncertainty with respect to all four intensity measurements can guard against this problem if the technical bias present in this case, e.g. the bias towards incorrectly calling T instead of A in later sequencing cycles is properly captured.4.