question about "plot_genes_in_pseudotime"

[vgettaa]

Dear all,
I tried to plot genes that changed expression over pseudotime and I got a plot that all dots are on the x-axis, the plot and code are as follow:
plot_genes_in_pseudotime(cds_order_1["CSN3",],color_cells_by="condition")

Then I added the "min_expr" mentioned in #602 . The plot seems to be incorrect since the curve is not fitted by the dots, the plot and code are as follow:

p1<-plot_genes_in_pseudotime(cds_order_1["CSN3",],color_cells_by="condition") 
p2<-plot_genes_in_pseudotime(cds_order_1["CSN3",],color_cells_by="condition",min_expr=min(p1$data$expectation)) 
p2

I checked the function plot_genes_in_pseudotime() https://rdrr.io/github/cole-trapnell-lab/monocle3/src/R/plotting.R. I think it seems that the problem in p1 was caused by round() function (cds_exprs <- reshape2::melt(round(as.matrix(cds_exprs))) line 1026), but I dont know why p2 would plot like that.

> table(p1$data$expression)
    0     1 
14598     1 
> table(p2$data$expression)
0.000180963278421048                    1 
               14598                    1 

Overall, I directly assigned the expression values based on line 1024-1025(https://rdrr.io/github/cole-trapnell-lab/monocle3/src/R/plotting.R). But I dont know whether the code is right or not:

a<-cds_order_1["CSN3",]
cds_exprs <- SingleCellExperiment::counts(a)
cds_exprs <- Matrix::t(counts(a))/size_factors(a)
p2$data$expression<-as.numeric(cds_exprs)
p2

Any suggestions are appreciated, Thanks!

[wbiangw]

Dear all, I tried to plot genes that changed expression over pseudotime and I got a plot that all dots are on the x-axis, the plot and code are as follow: plot_genes_in_pseudotime(cds_order_1["CSN3",],color_cells_by="condition")

Then I added the "min_expr" mentioned in #602 . The plot seems to be incorrect since the curve is not fitted by the dots, the plot and code are as follow:

p1<-plot_genes_in_pseudotime(cds_order_1["CSN3",],color_cells_by="condition") 
p2<-plot_genes_in_pseudotime(cds_order_1["CSN3",],color_cells_by="condition",min_expr=min(p1$data$expectation)) 
p2

I checked the function plot_genes_in_pseudotime() https://rdrr.io/github/cole-trapnell-lab/monocle3/src/R/plotting.R. I think it seems that the problem in p1 was caused by round() function (cds_exprs <- reshape2::melt(round(as.matrix(cds_exprs))) line 1026), but I dont know why p2 would plot like that.

> table(p1$data$expression)
    0     1 
14598     1 
> table(p2$data$expression)
0.000180963278421048                    1 
               14598                    1 

Overall, I directly assigned the expression values based on line 1024-1025(https://rdrr.io/github/cole-trapnell-lab/monocle3/src/R/plotting.R). But I dont know whether the code is right or not:

a<-cds_order_1["CSN3",]
cds_exprs <- SingleCellExperiment::counts(a)
cds_exprs <- Matrix::t(counts(a))/size_factors(a)
p2$data$expression<-as.numeric(cds_exprs)
p2

Any suggestions are appreciated, Thanks!

Thank you vgettaa for asking these questions while proposing the solution, I got the same question about why there's a round() function in source code and I'm glad to see you asking this —— but actually I have no clue about it yet.
I managed to get the proper plot by assigning the expression values just as you did, It looks really better.