Cannot join HTO libraries with different oligo sets

[RoganGrant]

I am trying to read multiple 10X libraries using the read10X() function. The libraries were run in two batches. Batch 1 was multiplexed with 4 samples per library (B0301 - B0304), whereas batch 2 was multiplexed with 5 samples per library (B0301 - B0305). I get the following error when I run the function:

Error: Matrices must have same number of rows in cbind2(argl[[i]], r)

I can easily re-run cellranger on batch 1 and just add B0305 and assume it will yield zero counts, however I can picture many cases where oligos will differ (and even overlap, but with different antibodies) between different batches for the same dataset. Alternatively, would it be possible to use a merge function instead of cbind2 which can generate NAs or zeros for non-overlapping rows.

Sorry, normally I would have a reprex but I don't know how to reasonably generate one here.

[mhkowalski]

Hello,

Thanks for your question. I would suggest that you use the read10X() function separately for each library. You should especially do this because I would run HTO demultiplexing separately for each library, and then merge the remaining singlets after demultiplexing. You can merge the two objects using the merge function (see here).

We'll work on creating a more informative error message here.

[RoganGrant]

This makes sense, thank you!