Figure 2

Estimated vs. actual template copy numbers, applying the same data to four methods of analysis. Symbols are mean values for each amplicon (n = 3, 10 samples each as in Figure 1; y-axis error bars, minimum and maximum copy number estimates; x-axis error bars, 10% s.d. from a priori error in pipetting and dilution). (a) Efficiency-independent method described in this paper (absolute quantification). RCAN1.1 dashed line (y = 1.05x - 5180, R² = 0.9996), β-actin dotted line (y = 1.00x - 83, R² = 0.9999). (b) Linear Regression of Efficiency (LRE) method (absolute quantification). Average correlations of fluorescence data to sigmoidal curves: R² = 0.9987 (n = 30, RCAN1.1, E_max = 0.912 ± 0.007 s.d.) and R² = 0.9999 (n = 30, β-actin, E_max = 0.941 ± 0.007 s.d.). RCAN1.1 dashed line (y = 0.618x - 4790, R² = 0.9903), β-actin dotted line (y = 0.240x - 5650, R² = 0.9751). (c) LinReg method (relative quantification). Y-axis is scaled to the geometric mean of data (open circle). Average correlations (log₁₀R _n vs. n, over 4 cycles) for determination of Φ: R² = 1.000 (n = 30, RCAN1.1) and R² = 1.000 (n = 30, β-actin). RCAN1.1 dashed line (y = 1.28x - 0.0951, R² = 0.9601), β-actin dotted line (y = 0.825x - 0.0761, R² = 0.9880). (d) Serial dilution method (relative quantification). Y-axis is scaled to the geometric mean of data (open circle). Average graph correlations (c _q vs. log₁₀A ₀ ) for determination of Ψ: R² = 0.9959 (RCAN1.1) and R² = 0.9947 (β-actin). RCAN1.1 dashed line (y = 1.19x - 0.261, R² = 0.9888), β-actin dotted line (y = 0.965x - 0.048, R² = 0.9820).