04-11-CellCycle.Rmd

# Cell cycle Assignment {#CellCycle}

In some datasets, the phase of cell cycle that a cell is in (G1/G2M/S) can account for 
alot of the observed transcriptomic variation. There may be clustering by phase, or 
separation in the UMAP by phase. 

Seurat provides a simple method for assigning cell cycle state to each cell. Other methods are available. 

More information about assigning cell cycle states to cells is in the [cell cycle vignette](https://satijalab.org/seurat/articles/cell_cycle_vignette.html)

```{r cc}
# A list of cell cycle markers, from Tirosh et al, 2015, is loaded with Seurat.  We can
# segregate this list into markers of G2/M phase and markers of S phase
s.genes   <- cc.genes$s.genes
g2m.genes <- cc.genes$g2m.genes

# Use those lists with the cell cycle scoring function in Seurat.
seurat_object <- CellCycleScoring(seurat_object, s.features = s.genes, g2m.features = g2m.genes)
```


Which adds S.Score, G2M.Score and Phase calls to the metadata.
```{r}
head(seurat_object@meta.data)
```


We can then check the cell phase on the UMAP. In this dataset, phase isn't driving the clustering, and would not require any further handling. 

```{r}
DimPlot(seurat_object, reduction = 'umap_harmony', group.by = "Phase")
```

Where a bias _is_ present, your course of action depends on the task at hand. It might involve 'regressing out' the cell cycle variation when scaling data `ScaleData(kang, vars.to.regress="Phase")`, omitting cell-cycle dominated clusters, or just accounting for it in your differential expression calculations.

If you are working with non-human data, you will need to convert these gene lists, or find new cell cycle associated genes in your species.