Cross-check analysis of the best linear regression genetic models. Polymorphism composition of each of the best models from Table 2 was applied to the four expression data sets and the -log10 P values for their fit are represented. Models defined as best in expression data sets with CEU LCL are shown in black (triangles for K8, squares for S and circles for C), and the best model defined with asthma cell lines in D is presented as white squares. Expression data sets labels in the X-axis are as in Table 2. Comparisons with the proper best model for each data set were either nonsignificantly inferior (n.s.) or inferior with *P < 0.05, **P < 0.01 or ***P < 0.001.