celda_irisapo

[ramikheireddine]

Hi, I am trying to load this load_all("~/gitProjects/celda_irisapo", recompile=T).

Can you please direct me to where i can download it. i cant seem to find the fix branch.
Thank you

[joshua-d-campbell]

Hi @ramikheireddine, this is the version of celda and decontX that we used for the paper:

https://github.com/campbio/celda/releases/tag/v0.7

You should be able to download and install this to reproduce the code in the paper. The basic algorithm of decontX hasn't changed. You may be able to install the latest version and get very similar results.

[ramikheireddine]

excuse me for the confusion as I am relatively new to this, I am supposed to load the celda file instead of celda-irisapo?
correct me if I'm wrong this is my script:
RNGkind(sample.kind = "Rounding")
library(devtools)
load_all("celda-0.7", recompile=T)

autoDecontX = function(sce, varGenes=5000, seed=12345, L=50, dbscan.eps=1, decontxIter=200) {
require(Matrix)
require(dbscan)
require(uwot)
require(scater)
require(scran)

sce = scater::normalizeSCE(sce)  # add the log2 normalized counts into sce object 

if( nrow(sce) <= varGenes) {
  topVariableGenes = 1:nrow(sce)
} else if( nrow(sce) > varGenes ) { 
  # Use the top most variable genes to do rough clustering (celda_CG & Umap & DBSCAN graph algorithm) 
  mvTrend = scran::trendVar(sce, use.spikes=FALSE) 
  decomposeTrend = scran::decomposeVar(sce, mvTrend) 
  topVariableGenes = order(decomposeTrend$bio, decreasing=TRUE)[1:varGenes]
} 
countsFiltered = as.matrix(counts(sce[topVariableGenes, ]))
storage.mode(countsFiltered) = "integer"

# Celda clustering using recursive module splitting
L = min(L, nrow(countsFiltered))    
	RNGkind(sample.kind = "Rounding")
	set.seed(12345)
initial.module.split = recursiveSplitModule(countsFiltered, initialL=L, maxL=L, perplexity=FALSE, verbose=FALSE)
initial.modules.model = subsetCeldaList(initial.module.split, list(L=L))
  fun_wd = getwd()
	cat("current working directory is \n", fun_wd)
  saveRDS(initial.module.split, file.path(fun_wd, "celda_obj.rds") )

# UMAP dimension reduction
	RNGkind(sample.kind = "Rounding")
	set.seed(12345)
fm = factorizeMatrix(countsFiltered, initial.modules.model, type="counts")
res.umap = uwot::umap(t(sqrt(fm$counts$cell)), n_neighbors=15, min_dist = 0.01, spread = 1)
reducedDim(sce, "QC_UMAP") = res.umap

# Use dbSCAN on the UMAP to identify broad cell types
totalClusters = 1
while(totalClusters <= 1 & dbscan.eps > 0) {
  res.dbscan = dbscan(res.umap, dbscan.eps)
  dbscan.eps = dbscan.eps - (0.25 * dbscan.eps)
  totalClusters = length(unique(res.dbscan$cluster))
}  

# If the counts is not matrix in the sce object, have to covnert the count into matrix to feed into celda_CG 
rm(countsFiltered)
counts = counts(sce) 
counts.classType = class(counts)     
counts = as.matrix(counts) 
storage.mode(counts) = "integer"    

if( totalClusters >=2) {
  # Perform decontamination with decontX 
  DecontXObject = decontX(counts=counts, z=res.dbscan$cluster, seed=seed, maxIter=decontxIter)
  DecontXObject$resList$estNativeCounts <- round(DecontXObject$resList$estNativeCounts) 

  if( counts.classType == "matrix") {
	  SummarizedExperiment::assay(sce, "decontX") <- DecontXObject$resList$estNativeCounts 
  } else if (counts.classType == "dgCMatrix") {
  SummarizedExperiment::assay(sce, "decontX") <- as(DecontXObject$resList$estNativeCounts, "dgCMatrix") 
  } else {
	  SummarizedExperiment::assay(sce, "decontX") <- DelayedArray::DelayedArray(DecontXObject$resList$estNativeCounts)
  }

  colData(sce)$DecontX_Contamination = DecontXObject$resList$estConp
  colData(sce)$DecontX_QuickCluster = res.dbscan$cluster	
} else {
  colData(sce)$DecontX_Contamination = 0
  colData(sce)$DecontX_QuickCluster = 1	    
}

return(sce) 

}

but im getting an error in '(function (command = NULL, args = character(), error_on_status = TRUE, ...':