Is it possible for a cell cluster to not have cells of a certain condition?
0
0
Entering edit mode
6 weeks ago

Hello, I'm new to seurat and bioinformatics in general. I've been doing data analysis on cells that have 5 different feeding conditions. After following the seurat pbmc tutorial, I made a violin plot showing the expression of a genetic marker, but I realized that one of my clusters only has cells for 4 of the conditions. There were almost 12k cells in total, so this did not make much sense to me as I thought all clusters would definitely have all conditions expressed. Is this possible?

Please help • 316 views
ADD COMMENT
0
Entering edit mode

Can you share your code and some basic QC plots, like a DimPlot of your samples split by condition, grouped by clusters, etc. It's difficult to give any advice without this information.

ADD REPLY
0
Entering edit mode

Okay so I used the raw data from NCBI GEO website and called it arc.data. Then, I created a seurat object

arc <- CreateSeuratObject(counts = arc.data, project = "arcuate median eminence", min.cells =3, min.features = 200, neames.field = 1, names.delim = '_')
ADD REPLY
0
Entering edit mode

Then I created some violin plots. The object has over 20k cells btw. I subsetted it just for neurons later, but these violin plots are for the entire thing. enter image description here

ADD REPLY
0
Entering edit mode

I followed the pbmc tutorial for cut-offs (used 5600 nfeatures), data normalization, scaling, pca, umap.

arc <- subset(arc, subset = nFeature_RNA > 200 & nFeature_RNA < 5600 & percent.mt < 20)

I hope I don't need to continue with all the other code because it's basically the pbmc tutorial in seurat. But after I finished with everything, I clustered the cells, and made a list of the neurons

neuron.list <- WhichCells(arc, idents = c(0,1,2,3,4))

After, I removed doublets with the doubletFinder function. I made an object by subsetting for the singlets:

#This is the object with all neurons (doublets and singlets). arc.original is just a copy of the original arc object I made at the start.
neurons_orig<-subset(arc.original, cells = neuron.list)

#Singlets
neurons_orig.singlets <- subset(neurons_orig, DF_hi.lo == "Singlet")

There were 11 current identiites: arc1, arc2, arc3, Chow, FemaleFasted, MaleFasted, FemaleFed, High Fat, Fasted, MaleFed, Refed. I renamed these with:

new.group.identities <- c("Control", "Control", "Control", "LFD", "Fasted", "Fasted", "Control", 
"HFD", "Fasted", "Control", "Refed")
names(new.group.identities) <- levels(neurons_orig.singlets)
neurons_orig<-RenameIdents(neurons_orig.singlets, new.group.identities)
levels(neurons_orig.singlets)

#I made a new metadata column for the identities, then i integrated data.
neurons_orig.singlets[["ids"]] <- Idents(object = neurons_orig.singlets)
neurons_int.list<-SplitObject(neurons_orig, split.by = "experiment.ids")
neurons_int.list <- lapply(X = neurons_int.list, FUN = function(x) {
    x <- NormalizeData(x)
    x <- FindVariableFeatures(x, selection.method = "vst", nfeatures = 2000)
})
neurons_int.features <- SelectIntegrationFeatures(object.list = neurons_int.list)
neurons.anchors <- FindIntegrationAnchors(object.list = neurons_int.list, anchor.features = neurons_int.features)
neurons.combined <- IntegrateData(anchorset = neurons.anchors)

Fast forward, I went through the cycle of redoing the pbmc tutorial (scale, pca, umap, etc.). I used 0.7 and 13 dimensions resolution to cluster.

I set the assay to "RNA", then I searched for conserved markers in each cluster. As a sample:

neuron.markers_0 <- FindConservedMarkers(neurons.combined, ident.1 = 0, grouping.var = ".ids", verbose = FALSE)

Now the problem is that when I did it with cluster 13, I got the following error:

neuron.marker_13 <- FindConserved Markers(neurons.combined, ident.1 = 13, grouping.var = "ids", verbose = FALSE)
#Error in ValidateCellGroups(object = object, cells.1 = cells.1, cells.2 = cells.2, : Cell group 1 has fewer than 3 cells

Also, after I subsetted for identity 13 (cluster 13) and made an object called cluster13.cells, I made a violin plot to see the distribution of a genetic marker in control ad high fat conditions, but this happens:

VlnPlot(cluster13.cells, features = "Ntrk2", idents = c("Control", "HFD"))

enter image description here

ADD REPLY

Login before adding your answer.

Traffic: 1146 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6