Clustering with monocle 3
I was trying to analyze a scRNA dataset (chromium 10x) using monocle 3.
I realized that changing the resolution parameter when applying modularity maximization did not change the capacity to discriminate subpopulations. The number of cells is 2k and UMAP visualization separates the cells into two regions, which are captured by Lovain modualrity.
When I changed different hyperparameters (PCs, resolution, …), I was not successful in obtaining an higher discrimination of the clusters. I can not find an explanation why I could not increase the number of clusters, like I would obtain using Seurat.
Do you have any suggestions?
• 26 views
Read more here: Source link