Differential expression between conditions between separate slingshot objects
nickhir opened this issue · 2 comments
I have single cell RNA seq data from 4 experiments. 2 are from WT mice, and two are from mutated mice. I am following this great tutorial how to use the package properly.
In the tutorial, you decide to fit one trajectory for the combined dataset (i.e. for both conditions), because the imbalance score is reasonable small. For my dataset, I observe very strong changes in celltype composition between mutant and WT and thus decided to split the dataset to contain only mutant or WT cells and the fitted trajectories separately.
After fitting the trajectories, I still discover similar linages, i.e. CelltypeA->CelltypeB->CelltypeC in both slingshot objects. I am now very interested in comparing these "similar" lineages to maybe identify genes that are different between the WT and mutant mice within this lineage.
My questions is if this is even possible or if you would recommend combining the dataset and fitting trajectories on the integrated data.
Any help is much appreciated!
Hi @nickhir ,
If the trajectory is similar but the composition changes it may still be reasonable to fit a single trajectory after proper integration across your samples/experiments. We have a workflow for cases like this, I recommend you to read the paper and hope that this helps you further. The condiments package also has several example case studies on the paper website provided by @HectorRDB.
I was not aware of this paper! Thank you very much, it helped tremendously!