Combining TCR and BCR

[msaezpalma]

Hi!
I'm currently working on TCR and BCR data and when I combine the Expression with the combineExpression function, both datsets are integrated in the Seurat Object metadata with the same name. That makes me not able to differentiate BCR and TCR in columns. I don't know if maybe I just need to add some line to the sunction or if I should continue doing all the analysis separately anthough I'll like to have all the data on one same Seurat Object.
Thanks a lot in advanced.
Maria

PD: The script I'm using now is this one:
bcr <- combineBCR(bcr_P7, samples = "L",
ID = "AV11081", threshold = 0.85)

tcr <- combineTCR(
tcr_P7,
samples = "L",
ID = "AV11081",
cells = "T-AB",
removeNA = FALSE,
removeMulti = FALSE,
filterMulti = FALSE
)

list.receptors <- c(tcr, bcr)

prova <- combineExpression(list.receptors,
P7_FNA,
cloneCall="gene",
chain = "both",
filterNA = FALSE,
cloneTypes = c(Rare = 1e-04, Small = 0.001, Medium = 0.01, Large = 0.1, Hyperexpanded = 1),
proportion = TRUE)

[ncborcherding]

Hey @msaezpalma,

Thanks for reaching out.

Could you use clusters to differentiate between TCR and BCR?

Alternatively, you could add a column to your meta.data using something like:

prova$receptor <- substring(prova$CTaa, 1,2)

This would extract the first 2 characters in the CTaa column to use to differentiate. Although that might not work for clones that have NA_XXXX. You might need to play around with it more to get all the info you need.

Thanks,
Nick

[msaezpalma]

Hi Nick,
Thanks for your fast answer. I did as suggested but my problem is that I do diferentiate the clusters of B cells and T cells but when I did the combineExpression with the tcr and bcr they end at the same columns.

So for example when doing the occupiedscRepertoire it does it with all the BCR and TCR data and I don't know how to ask only for the BCR or only for the TCR. With the line you gave me I still can not differenciate it.


I was trying now to rename the columns and combining the tcr first and then the bcr so that I get two different columns like CTaa_bcr/CTaa_tcr and then for doing the graphs changing again the name of the columns for bcr or tcr as I need to analyse.

looks now.
I hope I explained myself beacuse is a bit difficult.
Thanks a lot
Maria

[ncborcherding]

Hey Maria,

Unfortuantely, changing the column names will likely break the functions you want to use.

In scRepertoire v2 - most visualizations have an parameter called chain, which you could use to isolate on TCR or BCRs using "TRB" or "IGH" respectively.

You can install it with:

devtools::install_github("ncborcherding/scRepertoire")

There is a change in a lot of function names - so heads up if you are upgrading. The other issue here is that the 2 graphs you show (now clonalOverlay() and clonalOccupy()) are utilizing the precalculated variables from combineExpression() and do not have chain argument - so if you wanted those in particular the best way would be to subset your Seurat object into B and T cells.

Nick