How to add labels in the Karyo plot？

[sjasws]

Thank you for your outstanding tool! It is very useful in inferring CNV and is very accurate for my data.
Now I try to infer epithelial ovarian cancer cells (which are very closely associated with CNV) by epiAneufinder to distinguish normal ovarian epithelial cells. I mix all the cells and use epiAneufinder to infer the CNV, and then see which epithelial cells I have annotated are grouped together with non-epithelial cells, and these epithelial cells are considered normal epithelial cells, and those clustered alone with obvious CNV will be considered as tumor cells. Just as inferCNV done.
To facilitate monitoring, is there any way to add a legend of the cell type I already annotated to the plot_karyo_annotated () function, as the right part of your Figure 4a in Nat Commun 14, 5846 (2023)?
Thank you for your help!
With Best Regards
Lin

[thek71]

Hi Lin,

if I understand correctly, you have for each cell/barcode an annotation. What you can do is just load the annotation to R as a table with one column the cells and the second column the annotation, convert it to a data frame and use the plot_karyo_annotated () for creating a new plot, with the annotation of the cells as an extra to the original plot. In the vignette you can see an example and in your case the annot_dt data frame will be the annotation that you have.
I hope I understood correctly what you want to do. If not please correct me.

Best,
Katia

[sjasws]

Hi Lin,

if I understand correctly, you have for each cell/barcode an annotation. What you can do is just load the annotation to R as a table with one column the cells and the second column the annotation, convert it to a data frame and use the plot_karyo_annotated () for creating a new plot, with the annotation of the cells as an extra to the original plot. In the vignette you can see an example and in your case the annot_dt data frame will be the annotation that you have. I hope I understood correctly what you want to do. If not please correct me.

Best, Katia

Thank you for your reply.
I don't know if it could add the third column with cell type annotation? I add it but seem doesn't work.

[KatharinaSchmid]

Hi Lin,
I am not sure why you talk about a third column, the annotation data frame should have only two columns? To show the example for Katia's explanation from our vignette, you can use this function:

plot_karyo_annotated(res_table=res_table,
                     plot_path="epiAneufinder_results/karyo_annotated.png",
                     annot_dt=annot_dt)

The result table res_table you get after running the epiAneufinder main function and the annot_dt table is supposed to have two columns named cell and annot, like this:

> head(annot_dt)
                     cell  annot
1 cell-TCAGCTCTCGCGTTCT-1 Clone1
2 cell-TACGCAAAGTCGTGAG-1 Clone2
3 cell-TAACTTCTCCCGGGTA-1 Clone3
4 cell-ACCATCCGTTCTGAAC-1 Clone3
5 cell-CTCAGCTAGAGACTCG-1 Clone3
6 cell-ACAATCGTCTCATCCG-1 Clone4

Be aware that the cell column needs the barcodes exactly as in the epiAneufinder results table, which added a "cell-" prefix to each barcode.

Does this help you? And can you specify what doesn't work? You get an error message or the plot just looks weird?

[sjasws]

Dear @thek71 @KatharinaSchmid
Sorry, my English is very poor. I probably didn't express clearly. I draw a picture of what I want to achieve, that is, like the bottom panel, can an karyoplot generate multiple annotations for all cells at once?

[thek71]

Hi Lin,
the function works for one annotation.
But we can/will add this feature in the next release.
We will keep you posted when it's out.

Best regards,
Katia and Katharina

[sjasws]

Hi Lin, the function works for one annotation. But we can/will add this feature in the next release. We will keep you posted when it's out.

Best regards, Katia and Katharina

Dear Katia and Katharina,
Thank you for your answers and look forward to the next edition.
In addition, I have another question, in theory, will the CNV assessment of the same cell change with the total number of cells included in the analysis? In other words, will the results change if epiAneufinder analyze all the cells from different samples, versus if analyze each sample individually?
Because as far as I know, inferCNV on scRNA-seq will be different because it needs to set up a reference cell set.
Best Regards,
Lin

[thek71]

Hi Lin,

in epiAneufinder each cell is processed independently. There is no information flow from other cells. The CNVs predicted for a cell will be the same no matter how many other cells you process with it. If you use different parameters or different reference genome, that's when you will get different results.

[sjasws]

Hi Lin,

in epiAneufinder each cell is processed independently. There is no information flow from other cells. The CNVs predicted for a cell will be the same no matter how many other cells you process with it. If you use different parameters or different reference genome, that's when you will get different results.

Great, no wonder the results are so robust and accurate in my dataset!
All my questions have been answered perfectly. Thank you so much!
With Best Best Best Wishes,
Yours sincerely,
Lin

[Krithika-Bhuvan]

Hello. Wanted to follow up on this. Has the multi-annotation with karyoplot been implemented already, or is that in planning stages for the next version ?