search

cole-trapnell-lab / monocle-release

276 stars 116 forks source link

Why I can't find a ID column in my data and cds fvarLabels.? #268

Open Sophia409 opened 5 years ago

Sophia409 commented 5 years ago

@ctrapnell @rcannood @evolvedmicrobe @mrjc42 @eliah-o Why I can't find a ID column in my data and cds fvarLabels.?plot_genes_in_pseudotime functions relies on this column, but I can't find "id" .

> my_pseudotime_de %>% arrange(qval) %>% head()
  status           family pval qval gene_short_name num_cells_expressed use_for_ordering
1     OK negbinomial.size    0    0        Serpine2                2330             TRUE
2     OK negbinomial.size    0    0           Ramp1                1623             TRUE
3     OK negbinomial.size    0    0             Dbi                5190             TRUE
4     OK negbinomial.size    0    0          Pea15a                3707             TRUE
5     OK negbinomial.size    0    0          Atp1a2                1789             TRUE
6     OK negbinomial.size    0    0            Gatm                1587             TRUE
> tdcds
CellDataSet (storageMode: environment)
assayData: 54233 features, 5621 samples 
  element names: exprs 
protocolData: none
phenoData
  sampleNames: AAACCTGAGCAGCGTA AAACGGGCAGCTCGCA ... TGCTACCGTCGGCACT (5621 total)
  varLabels: orig.ident nCount_RNA ... State (15 total)
  varMetadata: labelDescription
featureData
  featureNames: 4933401J01Rik Gm26206 ... tdTomato (54233 total)
  fvarLabels: gene_short_name num_cells_expressed use_for_ordering
  fvarMetadata: labelDescription
experimentData: use 'experimentData(object)'
Annotation:

I follow this turtorial to do monocle analysis,in his picture,he selected the top 6 genes using "id", and there is a" id "among fvarLabels in his featureData,like this: default default Can you help me?

Xiaojieqiu commented 5 years ago

@Sophia409 the id column shown in the picture corresponds to the column storing the ensemble id of a gene. It is normally just the row name of the cell dataset (or the featureData). It seems like your cell dataset doesn't use ensemble id as row names. I suggest you to provide that information when you create the cell dataset. But you can still find the row name to subset the cell dataset for creating kinetic plots. For example using something like:
my_pseudotime_de %>% arrange(qval) %>% head() %>% row.names()

Sophia409 commented 5 years ago

@Xiaojieqiu Thank you for your advice, I use cell ranger to analysis my data and use the ensemble mouse reference. So I don't know why my cell dataset doesn't use ensemble id as row names. Another question, I need to merge my seurat data of 3 experimental time points with the Hippocampus data provided by Guided Clustering of the Mouse Cell Atlas: loom edition. I merged my own data by using the merge command and do trajectory analysis using monocle. But I don't know how to convert this loom file to seurat object or expression matrix, so that I can merge them with my data and do next analysis. The tutorial says that : Seurat offers a conversion function to go from Seurat objects to loom files. The reverse conversion is currently in progress. Does it means that I could't convert loom file to seurat? So what should I do to merge them? Can you give me some advice?Thank you very much!