Skip to main content
Fig. 1 | BMC Bioinformatics

Fig. 1

From: Nonlinear ridge regression improves cell-type-specific differential expression analysis

Fig. 1

Contour plot of the correlation coefficient between interaction terms \(W_{h} X_{k}\) and \(W_{{h^{\prime}}} X_{k}\). \(W_{h}\) and \(W_{{h^{\prime}}}\) represent proportions of cell types \(h\) and \(h^{\prime}\), and \(X_{k}\) represents the value of trait k. For this plot, we assume the coefficient of variation \({\text{CV}}\left[ {W_{h} } \right]\) and \({\text{CV}}\left[ {W_{{h^{\prime}}} } \right]\) to be equal. As the CV decreases 0.6, 0.4 to 0.2, the correlation coefficient raises > 0.5, > 0.7 to > 0.9, over most range of \({\text{Cor}}\left[ {W_{h} ,W_{{h^{\prime}}} } \right]\)

Back to article page
\