When subsetting a CCA aligned dataset, how should one renormalize/realign? #558
Labels
Comments
|
My question thought is does the data have to be realigned after this? It’s not as straight forward as the regular subletting a single seraut object with only one condition
|
|
Yes, you will need to reperform the alignment procedure. |
|
Alignment procedure finds variable genes in individual datasets. When you
subset an aligned dataset, the datasets are already combined. This is the
step I am confused about. Do I need to find variable genes independently in
each data set? I'm worried that the variable genes that would be used for
aligning would be heavily weighted by the cell types I'm not interested in.
Is there a standard way to find variable gens in that context before
aligning?
…On Thu, Jul 12, 2018 at 9:10 AM, satijalab ***@***.***> wrote:
Closed #558 <#558>.
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#558 (comment)>, or mute
the thread
<https://github.com/notifications/unsubscribe-auth/AYVPyJbI5oHMQDBhoLGa-rFetZFKAjL-ks5uF3USgaJpZM4UyhWl>
.
--
Alex Chitsazan
Research Scientist/Engineer
Wills Lab | Department of Biochemistry
University of Washington
503-916-9064
|
mojaveazure
pushed a commit
that referenced
this issue
May 28, 2021
check reference.reduction in FindTransferAnchors
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I have a CCA dataset that was aligned given three developmental timepoints (3 samples). Many cell types within this CCA aligned dataset are not what we are interested in and are hoping to do a similar analysis on only a couple clusters of interest (from all 3 samples). When one subsets the data to isolate clusters of interest, how should the data be realigned? The way that variable genes are found within each dataset is different prior to alignment. Is there a standard way of how one should approach looking at/normalizing a subset of a CCA aligned dataset?
The text was updated successfully, but these errors were encountered: