Studying the Complex Expression Dependences between Sets of Coexpressed Genes
Figure 3
Four expression relationships are shown. The different sample conditions of the expression matrix (the sample series) constitute the data cloud. The PCOP describes the expression-relationship inner pattern. (b) and (c) show coexpressed genes. HLA-A and HLA-F are coexpressed genes (b) and GRAMD1A and NREP are also coexpressed genes (c). (a) and (d) show nonlinear expression relationships of type, an activation relationship. Since HLA-F and GRAMD1A maintain a nonlinear expression relationship of type (a), HLA-F is coexpressed with HLA-A (b), and GRAMD1A is coexpressed with NREP (c), therefore HLA-A and NREP (C5orf13) maintain a nonlinear expression relationship of the same type (d). HLA-A and GRAMD1A would also maintain a nonlinear relationship of type , and HLA-F and NREP would maintain a nonlinear relationship of the same type. This is the key point of our approach: all the nonlinear expression relationships that relate genes from two sets of coexpressed genes will have the same type of curve.
