niaid / baseline Goto Github PK

Hi @kotliary ,

I downloaded the file H1_day0_demultilexed_singlets.RDS and I'm trying to follow your code to build my own SingleCellExperiment, but I'm a bit confused of which rows are the surface markers and which are genes:

You start by loading and splitting by batch,

# read in final singlets make list of seurat objects indexed by batch. 
h1 = readRDS(file = "data/H1_day0_demultilexed_singlets.RDS") %>% SetAllIdent(id = "batch")
h1b1 = SubsetData(h1, ident.use = "1")
h1b2 = SubsetData(h1, ident.use = "2")

Then you use h1b1 and h1b2 as your positive ADT counts

# make list of positive protein matrices by batch 
stained = list(h1b1, h1b2)
pos_adt = lapply(stained, function(x){x@[email protected]})

then you use the same h1b1 and h1b2 to analyze the RNA data

sc = list(h1b1, h1b2)
sc =  lapply(sc, function(x){ Convert(from = x, to = "sce") })  
suppressMessages(library(scater))
sc = lapply(sc, FUN = calculateQCMetrics)

I don't understand what part of the seurat object has the RNA and what part the ADT.

Thank you,
Stephany

Hi, your paper is really great!
i'm analysing some data generated over 3 channels/lanes of a 10x Chip. I am normalizing each channel-worth of ADT data with your DSB package, as the background is channel-specific. I need to merge these 3 channels in one dataset afterwards. how can i address the batch effect that the 3 different normalizations would probably introduce? in the paper you mention using limma

In addition, to account for droplet-to-droplet differences in the ADT capture rate as well as background noise from unbound antibody, we denoised each cell by removing a covariate corresponding to the background counts for each cell using the removeBatchEffect function in limma

can you clarify when and how this normalization is applied? before or after merging your channels (i suppose your data is over 6 channels from the methods but I'm unsure).

thank you!

Dear @kotliary

I'm trying to run the provided workflow for CITE-seq. I faced an error since you've used SetAllIdent in here. But, this function is removed in the new version of Seurat (v3) as mentioned here.

I've also tried to install the previous version of Seurat. But it needs SDMTools package which is removed from CRAN because of its dependency adehabitat .

I was wondering if you can help me to change this line of code (and also line 18/21) to the alternative function from Seurat version 3 called Idents .

Regards,
Sina