By utilizing the scRNA-seq data coming from mammary epithelial cells of young and old cells, I shortlisted the aging signature genes (13 genes) by performing the differential gene expression (DGE) analysis on the scRNA-seq data. Then I utilized this13 genes (aging signatures) scRNA-seq data to construct an LDA model to predict the age of cells. It's straightforward that I can use this model to predict the age of the cells with scRNA-seq data without the age information. However, I want to use this same model to predict the age of the cells that are sequenced with BULK-RNA seq data. Can we predict the age of the cells with BULK-RNA seq data using the same LDA model trained on scRNA-seq data? Kindly explain the reason in either of the cases (Yes or NO). Also, If yes, then which strategy works best for making the Bulk-RNA data comparable to use it for age prediction in the LDA model?
I cannot give a precise answer, as ML is not my field, but generally, why don't you apply the method to pseudobulks of the exact single-cell data that you used? That will give a feel for how it performs at best with the same data but "bulked".
Thank you so much for this insight. I just want to clarify another point here. I have used scRNA seq data from mice mammary epithelial cells with young and old cells. I have used BUlk RNA seq from human mammary epithelial cells from young and old cells. I performed DGE analysis on scRNA seq data using young epithelial cells as reference. Likewise, DGE is performed on BULK data while setting young samples as reference. I was expecting that since the data for scRNA and bulk RNA came from the same tissue(mammary epithelial cells) so DEGs from scRNA will also differentially expressed in BULK data. However, the DEGS obtained from scRNA are not common to BUlK DEGs. Can you kindly shed some light on this?
I cannot comment here. You are asking why results between two experiments are different. I do not know without seeing the data.
ATpoint thanks you for your timely response. Can someone else shed light if we can compare the trend of DEGs using the scRNA and Bulk RNA data taken from the same tissue (mammary epithelial cells)?