implementation of seurat DotPlot function #882

NoemieL · 2021-07-12T22:17:29Z

NoemieL
Jul 12, 2021

Hi,
Could it be possible to have a similar function as the one proposed in seurat with DotPlot (https://satijalab.org/seurat/archive/v3.0/visualization_vignette.html) that provide a graph with the percentage of cells expressing the feature and the average expression level in each cluster. Could be nice to have this available for GeneScore and GeneIntegration.

Thanks

sylestiel · 2021-12-09T17:44:07Z

sylestiel
Dec 9, 2021

Were you able to figure this out?

2 replies

NoemieL Dec 16, 2021
Author

Yes, you could find my script here https://github.com/NoemieL/bubble-plot-ArchR
Please let me know if something is not clear.

ZayaAyush Feb 23, 2024

Thank you, NoemieL. Can you please tell me how to change the color in a bubble plot?

sylestiel · 2021-12-16T22:07:52Z

sylestiel
Dec 16, 2021

Thanks NoemieL for the script.

I have a few questions:
How did you generate the gene_list?

Also can the same be used to plot using the gene score matrix? If so how should I modify the script.

5 replies

NoemieL Dec 16, 2021
Author

For gene_list, you could do gene_list=c("TP53","BCL2","TGFB") or take the list of gene from ArchR

To use the gene score : genescore= getMatrixFromProject(ArchRProj, useMatrix="GeneScoreMatrix")

sylestiel Dec 16, 2021

Thanks for your prompt reply.

matteozoia4 Nov 22, 2022

Dear Noemie many thanks for your script!

I encounter an error during bubble dataframe creation based on GeneExpressionMatrix.

Dataframe creation :

bubble_plot_info=data.frame()
for(i in gene_list){
  for(k in 1:length(unique(geneinteinfo2$Clusters))){
    a=nrow(bubble_plot_info)
    l=unique(geneinteinfo2$Clusters)[k]
    bubble_plot_info[a+1,"gene_name"]=i
    bubble_plot_info[a+1,"Clusters"]=l
    eval(parse(text=(paste("bubble_plot_info[",a,"+1,'pct_exp']=(length(geneinteinfo2[(geneinteinfo2$",i,">0 & geneinteinfo2$Clusters=='",l,"'),'",i,"'])/nrow(geneinteinfo2[geneinteinfo2$Clusters=='",l,"',]))*100", sep=""))))
    eval(parse(text=(paste("bubble_plot_info[",a,"+1,'avg_exp']=mean(geneinteinfo2[geneinteinfo2$",i,">0 & geneinteinfo2$Clusters=='",l,"','",i,"'])", sep=""))))
  }
}

Error :

In mean.default(geneinteinfo2[geneinteinfo2$HCN4 > 0 &  ... :
  argument is not numeric or logical: returning NA

How could I overcome this error?

Many thanks for he help!

MZ

NoemieL Nov 22, 2022
Author

Hi Matteo,
Does it work with other genes? Are you sure your HCN4 gene is expressed in your cells? Or is it present in your GeneExpressionMatrix?
If not, your values may not be numerical. It's hard to tell since you didn't show how your data is. Could you give more information about what you have done before and how your data looks like?
Thanks

NoemieL Nov 22, 2022
Author

I looked in my data and your gene is not present in the GeneExpressionMatrix, I also tried the aliases BRGDA8, EIG18, and SSS2. I think this is the problem and the only way to get it is to find a way to not exclude its region on your scRNAseq data.

sylestiel · 2021-12-17T17:06:34Z

sylestiel
Dec 17, 2021

Any suggestions on how to specify the order in which the genes are plotted?

Also how to are used the predicted group names of clusters post integration?

7 replies

sylestiel Mar 8, 2022

@NoemieL

How can I modify the script to incorporate scaled values instead of avg_expression?

NoemieL Mar 8, 2022
Author

I am not sure I understand what you want to do. If you want to scaled the avg_expression you could create a new column in geneinteinfo2 scale(geneinteinfo2$avg_expression) and then select this column as x in the ggplot.

sylestiel Mar 10, 2022

You are understanding me correctly. However, I'm not familiar with scripting. So don't know how to incorporate scaling on average expression for the dot or bubble plot.
I'm getting the following error:

colnames(geneinteinfo2)=scale(geneinteinfo2$avg_expression)
Error in array(x, c(length(x), 1L), if (!is.null(names(x))) list(names(x), :
'data' must be of a vector type, was 'NULL'

NoemieL Mar 11, 2022
Author

Sorry, it is on bubble_plot_info and not on geneinteinfo2 that you need to add a column.
After creating your bubble_plot_info, do:
bubble_plot_info[is.na(bubble_plot_info$avg_exp),"avg_exp"]=0 #to replace NA value by 0

for(i in gene_list){
bubble_plot_info[which(bubble_plot_info$gene_name==i),'scale(avg_exp)']=scale(bubble_plot_info[which(bubble_plot_info$gene_name==i),'avg_exp'])
} # to scale our data for each gene

and change in the ggplot color = "scale(avg_exp)"

sylestiel Mar 12, 2022

Appreciate it!! It worked.

sylestiel · 2022-03-15T15:39:02Z

sylestiel
Mar 15, 2022

@NoemieL

I would like to change the script for percent expression to reflect percent of cells in all clusters and not just in individual clusters. How may I incorporate that into your script for dot plots?

2 replies

NoemieL Mar 29, 2022
Author

Sorry I was busy. If you want the percentage of cells that expressed the gene without taking into account the clustering, you just have to remove this geneinteinfo2$Clusters=='",l,"' that is the condition of being in a specific cluster.

I would do another data.frame as follow:

bubble_plot_info=data.frame()
for(i in gene_list){
a=nrow(bubble_plot_info)
bubble_plot_info[a+1,"gene_name"]=i
eval(parse(text=(paste("bubble_plot_info[",a,"+1,'pct_exp_total']=(length(geneinteinfo2[(geneinteinfo2$",i,">0 ,'",i,"'])/nrow(geneinteinfo2))*100", sep=""))))
eval(parse(text=(paste("bubble_plot_info[",a,"+1,'avg_exp']=mean(geneinteinfo2[geneinteinfo2$",i,">0 ,'",i,"'])", sep=""))))
}

sylestiel Mar 29, 2022

Thank you!!!

matteozoia4 · 2022-11-23T09:14:47Z

matteozoia4
Nov 23, 2022

Dear Noemie, many thanks for your tempestive answer! I am looking through this, I may indeed have set accidentally to uppercases the gene names, since working with mouse this could be definitely the issue for matching. As well I need to employ a matrix which has not been generated through Integration of RNA-ATAC, but following 10x RNA/ATAC pipeline. I therefore miss the integration matrix and so far I haven't got the chance to check for eventual discrepancies in structure or naming with previous matrices (such as GeneExpression or GeneScore ones). I am on this will be back to you soon! Many thanks again for your help! MZ Matteo Zoia, Ph.D. Student Gene Regulation & Cardiac Development Osterwalder Lab Department for BioMedical Research (DBMR) University of Bern Murtenstrasse 28, 369 CH - 3008 Bern SWITZERLAND

…

________________________________ Da: NoemieL ***@***.***> Inviato: martedì, 22. novembre 2022 18:09:53 A: GreenleafLab/ArchR Cc: Zoia, Matteo (DBMR); Comment Oggetto: Re: [GreenleafLab/ArchR] implementation of seurat DotPlot function (Discussion #882) I looked in my data and your gene is not present in the GeneExpressionMatrix, I also tried the aliases BRGDA8, EIG18, and SSS2. I think this is the problem and the only way to get it is to find a way to not exclude its region on your scRNAseq data. — Reply to this email directly, view it on GitHub<#882 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AUBHT3FLUIJEHMFZMGNIZ2TWJT46DANCNFSM5AHXFLBQ>. You are receiving this because you commented.Message ID: ***@***.***>

0 replies

matteozoia4 · 2022-11-23T15:09:22Z

matteozoia4
Nov 23, 2022

Dear Noemie, In the end the uppercase syntaxis was the issue, my bad I didn't looked at this in the immediate. Many thanks for your time and additional explanations. MZ Matteo Zoia, Ph.D. Student Gene Regulation & Cardiac Development Osterwalder Lab Department for BioMedical Research (DBMR) University of Bern Murtenstrasse 28, 369 CH - 3008 Bern SWITZERLAND

…

________________________________ Da: Zoia, Matteo (DBMR) Inviato: mercoledì, 23. novembre 2022 10:14:40 A: GreenleafLab/ArchR; GreenleafLab/ArchR Cc: Comment Oggetto: Re: [GreenleafLab/ArchR] implementation of seurat DotPlot function (Discussion #882) Dear Noemie, many thanks for your tempestive answer! I am looking through this, I may indeed have set accidentally to uppercases the gene names, since working with mouse this could be definitely the issue for matching. As well I need to employ a matrix which has not been generated through Integration of RNA-ATAC, but following 10x RNA/ATAC pipeline. I therefore miss the integration matrix and so far I haven't got the chance to check for eventual discrepancies in structure or naming with previous matrices (such as GeneExpression or GeneScore ones). I am on this will be back to you soon! Many thanks again for your help! MZ Matteo Zoia, Ph.D. Student Gene Regulation & Cardiac Development Osterwalder Lab Department for BioMedical Research (DBMR) University of Bern Murtenstrasse 28, 369 CH - 3008 Bern SWITZERLAND

________________________________ Da: NoemieL ***@***.***> Inviato: martedì, 22. novembre 2022 18:09:53 A: GreenleafLab/ArchR Cc: Zoia, Matteo (DBMR); Comment Oggetto: Re: [GreenleafLab/ArchR] implementation of seurat DotPlot function (Discussion #882) I looked in my data and your gene is not present in the GeneExpressionMatrix, I also tried the aliases BRGDA8, EIG18, and SSS2. I think this is the problem and the only way to get it is to find a way to not exclude its region on your scRNAseq data. — Reply to this email directly, view it on GitHub<#882 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AUBHT3FLUIJEHMFZMGNIZ2TWJT46DANCNFSM5AHXFLBQ>. You are receiving this because you commented.Message ID: ***@***.***>

0 replies

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

implementation of seurat DotPlot function #882

{{title}}

Replies: 6 comments 16 replies

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

Select a reply

implementation of seurat DotPlot function #882

Replies: 6 comments · 16 replies

NoemieL Dec 16, 2021 Author

NoemieL Dec 16, 2021 Author

NoemieL Nov 22, 2022 Author

NoemieL Nov 22, 2022 Author

NoemieL Mar 8, 2022 Author

NoemieL Mar 11, 2022 Author

NoemieL Mar 29, 2022 Author

Replies: 6 comments 16 replies

NoemieL Dec 16, 2021
Author

NoemieL Dec 16, 2021
Author

NoemieL Nov 22, 2022
Author

NoemieL Nov 22, 2022
Author

NoemieL Mar 8, 2022
Author

NoemieL Mar 11, 2022
Author

NoemieL Mar 29, 2022
Author