I am having a single cell RNA seq data from ~500 Cells. I do understand this number is very low for present day analysis tools like Seurat. So, I want to know what if one have a such a small number of cells which tool works best?
Thanks
$\begingroup$
$\endgroup$
5
-
$\begingroup$ you can perfectly use seurat as part of your analysis pipeline. It is hard to suggest tools or statistics without knowing the type of data you are looking at and what questions you want to address. $\endgroup$– Mack123456Jul 26, 2019 at 21:59
-
$\begingroup$ @Mack123456 Thanks for your comment. I am trying to use single cell RNA seq data on MiNION. On a trial basis I just started with ~500 Cells. $\endgroup$– user3377241Jul 26, 2019 at 22:09
-
$\begingroup$ Hi, you will need to forward as much information as possible, or simply accept @Mark123456 answer. The problem is that we don't know the objective of the study. $\endgroup$– M__ ♦Jul 27, 2019 at 12:16
-
$\begingroup$ @MichaelG. Thanks. I am ready to accept the comment as answer. But I cant find the accept answer tag in comment section. $\endgroup$– user3377241Jul 27, 2019 at 16:30
-
$\begingroup$ Hi @user3377241 no need these are just comments not a full answer. If you look at the other questions it's detailed technical advice rather than broad answers. If you have a specific technical issue you can return and ask about it $\endgroup$– M__ ♦Jul 27, 2019 at 16:40
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
You can perfectly use any of the single cell processing packages out there. Seurat, Monocle3, Scanpy, Kallisto, etc. Following some of the tutorials and vignettes will most likely help you figuring out how to proceed.
These packages provide useful functions to explore single cell data, process the data and each of them has a series of advanced features to address specific questions. You don't need 10k cells to look at number of gene/cell, transcript/cells, % mitochondrial genes, etc. You can perfectly normalize and standardize 500 cells and run PCA, and analyze the complexity that describes your data.
To address if 500 cells is enough for you pilot we need to know what question you want to address. The following tool for example gives some intuition on how many cells to sample if you wish to detect n-cells per cluster.
https://satijalab.org/howmanycells
Hopes this helps
