search

AntonioDeFalco / SCEVAN

R package that automatically classifies the cells in the scRNA data by segregating non-malignant cells of tumor microenviroment from the malignant cells. It also infers the copy number profile of malignant cells, identifies subclonal structures and analyses the specific and shared alterations of each subpopulation.
https://www.nature.com/articles/s41467-023-36790-9
GNU General Public License v3.0
92 stars 25 forks source link

The fibroblast, EC, and myofibroblasts were also identified as tumor. What are the reasons for this result? #60

Open gloriafight opened 1 year ago

gloriafight commented 1 year ago

I'm using SCEVAN to analyze a tumor sample. However, when I use the total cells as input. The result is very strange. image image image

The marker is shown in the figure. We can know that C4 is EC, C11 is fibroblast, C0 and C10 is Epithelial cell. I want to distinguish normal cells and malignancy cells in epithelial cells. The code is as follows: `count_mtx = cr@assays$RNA@counts

results <- pipelineCNA(count_mtx,sample = "aaaa", par_cores = 8)`

AntonioDeFalco commented 1 year ago

Is it possible to share the aaaaheatmap.png file generated by SCEVAN? Are you analyzing human or mouse sample?

gloriafight commented 1 year ago

Is it possible to share the aaaaheatmap.png file generated by SCEVAN? Are you analyzing human or mouse sample? This is the heatmap. image

It is human sample.

AntonioDeFalco commented 1 year ago

The separation obtained by copy number alteration seems quite correct (although because of the perhaps not very high sample quality the heatmap is not so clear). The classified malignant cells do indeed seem to have obvious common alterations on some chromosome types 1, 3, 7 and 10. What kind of tumor is it? In case you are sure that those cells are not cancer you can pass to SCEVAN all the cells that you think are confidently normal as parameter _normcells to get a better classification.

Regards.

gloriafight commented 1 year ago

The separation obtained by copy number alteration seems quite correct (although because of the perhaps not very high sample quality the heatmap is not so clear). The classified malignant cells do indeed seem to have obvious common alterations on some chromosome types 1, 3, 7 and 10. What kind of tumor is it? In case you are sure that those cells are not cancer you can pass to SCEVAN all the cells that you think are confidently normal as parameter _normcells to get a better classification.

Regards.

Thank you for your suggestion. The heatmap certainly looks like it. It is lung cancer. I can designate these cells as normal cells to try to define malignant cell. However, does this indicate that some normal cell types may also have aneuploid at certain sites on the chromosome. Or it is because of the 10X scRNAseq data. Or is this something else? This is really interesting.