correct batch effect for data from different organ and integrate multiple studies

[Beki-seq]

Good afternoon,

I hope you can give me some suggestions regarding my research.

I intend to use Harmony to clean batch effects in my dataset, which includes multiple organs from different mice. I applied Harmony to clean the batch effect with the following code:

"library(harmony) combine_datalist <- SCTransform(combine_datalist) %>% RunPCA() combine_datalist <- RunHarmony(combine_datalist, group.by.vars="organ", assay.use="SCT", max.iter.harmony=40, max.iter.cluster=40, kmeans_init_nstart=20, kmeans_init_iter_max=100)"

However, the results were not ideal, as one cluster of liver cells crossed with one cluster of kidney cells. Could you please give me some suggestions on how to use your package to clean batch effects in my dataset?

Best, Andy

[ahdee]

Not the developer but for one thing it would not be too suprising to see some cross contimation as an "organ" is not a pure cell line and you would expect certain cell type to be shared between various organs. I would suggest to use some downstream analysis to annotate celltype and see what type of cells are the one overlapping? perhas just some common blood cell or fibroblast? anyway just a thought.