Figure 2.

Incorrect use of C_q for removal of between-plate variation. Some methods to remove between-plate variation are erroneously based on dividing the C_q values on a plate by a factor per plate. The graph is based on hypothetical samples with three true expression levels (N₀ values = 10, 100 and 1000) measured on nine plates with between-plate differences ranging from 0.33 times down to 3 times up. The measured expression levels were converted into C_q values, divided by the factor per plate and re-converted into “corrected” expression levels. Between-plate correction should preserve the expression ratios on each of the plates in the experiment. Therefore, the 100- and 10-fold ratio between the highest and the middle N₀ (circle) and the highest and the lowest N₀ (triangle), should not be preserved. The graph shows that dividing C_q values will lead to distortion of expression ratios.