Doctor G

In the end, I went to FastMNN itself (as you suggest) and constructed the object directly rather than through the Seurat wrapper. It was a bit annoying, but worked well...

Oh, I think I know this one. For reasons I do not understand, the row order is different (and possibly the resulting rowData) on the SCE that come in via...

And if you set `addGeneScoreMat = FALSE` the arrow file appears to be created successfully (though the [empty] tmp directory remains). So there does indeed seem to be something "special"...

(Facepalm) I’m surprised that list error made it through the TxDb construction. Thanks for the catch. Do you expect any issues due to some genes being close to chromosome/scaffold ends...

> There have been other instances where genes within 2 kb of scaffold ends had to be excluded from the transcript database (#288 Great, thanks for the tip. So far...

The status of this error seems to be solved (not getting the error in the current ArchR version), but rather getting warnings that can be ignored.

Note: Changing `if (nFrags == 0)` to `if (nFrags < 2)` in `.getFragsFromArrow` does indeed seem to resolve the issue. Please let me know if this was an unintentionally bad...

Great, thanks! My modified code does what I need locally, but the error is part of a general theme -- for non-model species with small scaffolds (or small chromosomes): 1)...

Actually, looking at the inputs, the issue may be upstream -- the file `clusters_tmp.tsv` is empty and the log says ``` total loci used 23808 thread 6 iteration 0 done...

Ref ``` Singularity> head /gstore/scratch/u/garfield/souporcell//SAM24416638/ref.mtx %%MatrixMarket matrix coordinate real general % written by sprs 172837 1301589 9267869 1 91640 0 1 173540 1 1 263450 1 1 297056 1 1...