Problems with analyzing fluorescence data in ISH datasets

I am using the schizophrenia ISH dataset. The problem is that the image has three stains. According to the ISH Whitepaper, red is for high expression, green for medium, and blue for low, and each cell contains all three. I discovered that any cell with a high expression value has a strong red. Therefore, I wonder if the red stain intensity value can roughly represent the expression value of the cell. If not, please inform me about a method to combine the intensity data of the three stains.

Thank you!

Hello @Harry! The schizophrenia dataset is colormetric ISH of a single gene per tissue slice and the express mask is a false color heat map of the grayscale expression intensity level of that gene. You can find more information about the automated gene expression detection methods in (Lein et al., 2007; Supplemental Methods 2).