convert monolce3 into GeneSwitches

[junjunjing]

Hello,
I am new to GeneSwitches and I just started to learn how to import monocle3 instead of monocle2 as input for GeneSwitches. Is there a way to archive this? Thanks for your help.
Junjun

[seqyuan]

我是这样做的：

# rds 为seurat对象，cds为monolce3对象
rds@meta.data$pseudotime <- cds@principal_graph_aux$UMAP$pseudotime
Seurat2GSsce <- function(Seuratrds, clusters){
    Seuratrds <- subset(Seuratrds, seurat_clusters %in% clusters)
    sce <- SingleCellExperiment(assays = List(expdata = Seuratrds@assays$RNA@data))
    colData(sce)$Pseudotime <- Seuratrds@meta.data$pseudotime
    UMAP <- Seuratrds@reductions$umap@cell.embeddings
    reducedDims(sce) <- SimpleList(UMAP = UMAP)
    return (sce)
}

# 根据monocle3的轨迹判断有两条发育路线
sce_p1 <- Seurat2GSsce(rds, c(11,4,13,10,2,1))
sce_p2 <- Seurat2GSsce(rds, c(11,4,13,12,15,6,5))

sce_p1 <- binarize_exp(sce_p1, ncores = 2)
sce_p2 <- binarize_exp(sce_p2, ncores = 2)
# 后面可以比较 sce_p1和sce_p2

但是我遇到的问题是binarize_exp运行步骤 总是被 Killed 任务退出，或者出现下面的错误

WARNING! NOT CONVERGENT! 
number of iterations= 1000 
Error in oupBinary$mu2 - oupBinary$mu1 : 
  non-numeric argument to binary operator
Calls: binarize_exp -> [ -> [.data.frame
In addition: Warning message:
In mclapply(rownames(LogCountsadd), function(iGene) { :
  scheduled core 1 encountered error in user code, all values of the job will be affected
Execution halted

[pushpinder-bu]

@junjunjing You found a way how to import monocle3 object?

[Jiayi-Zheng]

I tried the code from seqyuan and encountered:

> sce <- binarize_exp(sce, ncores = 2)
Error in `*tmp*`[oupBinary$lambda1 < 0.1, ] : 
  incorrect number of dimensions
In addition: Warning messages:
1: In mclapply(rownames(LogCountsadd), function(iGene) { :
  all scheduled cores encountered errors in user code
2: In oupBinary$passBinary <- TRUE : Coercing LHS to a list

[SheldonXu1999]

I tried the code from seqyuan and encountered:

> sce <- binarize_exp(sce, ncores = 2)
Error in `*tmp*`[oupBinary$lambda1 < 0.1, ] : 
  incorrect number of dimensions
In addition: Warning messages:
1: In mclapply(rownames(LogCountsadd), function(iGene) { :
  all scheduled cores encountered errors in user code
2: In oupBinary$passBinary <- TRUE : Coercing LHS to a list

I used to have the same problem, and setting the "fix_cutoff" parameter TRUE works for me.
sce_p1 <- binarize_exp(sce_p1, fix_cutoff = TRUE, binarize_cutoff = 0.2)

[SheldonXu1999]

我是这样做的：

# rds 为seurat对象，cds为monolce3对象
rds@meta.data$pseudotime <- cds@principal_graph_aux$UMAP$pseudotime
Seurat2GSsce <- function(Seuratrds, clusters){
    Seuratrds <- subset(Seuratrds, seurat_clusters %in% clusters)
    sce <- SingleCellExperiment(assays = List(expdata = Seuratrds@assays$RNA@data))
    colData(sce)$Pseudotime <- Seuratrds@meta.data$pseudotime
    UMAP <- Seuratrds@reductions$umap@cell.embeddings
    reducedDims(sce) <- SimpleList(UMAP = UMAP)
    return (sce)
}

# 根据monocle3的轨迹判断有两条发育路线
sce_p1 <- Seurat2GSsce(rds, c(11,4,13,10,2,1))
sce_p2 <- Seurat2GSsce(rds, c(11,4,13,12,15,6,5))

sce_p1 <- binarize_exp(sce_p1, ncores = 2)
sce_p2 <- binarize_exp(sce_p2, ncores = 2)
# 后面可以比较 sce_p1和sce_p2

但是我遇到的问题是binarize_exp运行步骤 总是被 Killed 任务退出，或者出现下面的错误

WARNING! NOT CONVERGENT! 
number of iterations= 1000 
Error in oupBinary$mu2 - oupBinary$mu1 : 
  non-numeric argument to binary operator
Calls: binarize_exp -> [ -> [.data.frame
In addition: Warning message:
In mclapply(rownames(LogCountsadd), function(iGene) { :
  scheduled core 1 encountered error in user code, all values of the job will be affected
Execution halted

I used to have the same problem shows
Error in oupBinary$mu2 - oupBinary$mu1 : non-numeric argument to binary operator,
and setting the "fix_cutoff" parameter TRUE works for me.
sce_p1 <- binarize_exp(sce_p1, fix_cutoff = TRUE, binarize_cutoff = 0.2)