Issues about scCombination

[Tianqi-Ma]

Hi,

I was using scCombination to do an integrated analysis on 3 10x datasets. (scStatistics and scAnnotation had no errors on these 3 datasets individually and results were all achieved.)

When I set the integration to "Harmony", it returned an error message like this:

[2022-01-07 08:31:58] START: RUN ScCombination [2022-01-07 08:31:58] -----: sample data combination [1] "BHP006" [1] "BHP012" [1] "B211635C1L1_211215" [2022-01-07 08:32:23] -----: combine data by Harmony Warning: Non-unique cell names (colnames) present in the input matrix, making unique Error in assay.use %||% DefaultAssay(object) : could not find function "%||%" Calls: runScCombination -> %>% -> RunHarmony -> RunHarmony.Seurat Execution halted

Then I changed the integration method to '"NormalMNN", it also returned an error message after finding anchors, like this:

[2022-01-07 08:45:35] START: RUN ScCombination [2022-01-07 08:45:35] -----: sample data combination [1] "BHP006" [1] "BHP012" [1] "B211635C1L1_211215" [2022-01-07 08:45:53] -----: combine data by normal cell MNN Computing 2000 integration features Scaling features for provided objects Finding all pairwise anchors Running CCA Merging objects Finding neighborhoods Finding anchors Found 23550 anchors Filtering anchors Retained 9245 anchors Running CCA Merging objects Finding neighborhoods Finding anchors Found 20682 anchors Filtering anchors Retained 8171 anchors Running CCA Merging objects Finding neighborhoods Finding anchors Found 25783 anchors Filtering anchors Retained 13653 anchors Error in idx[i, ] <- res[[i]][[1]] : number of items to replace is not a multiple of replacement length Calls: runScCombination ... FindWeights -> NNHelper -> do.call -> AnnoyNN -> AnnoySearch Execution halted

I just wonder if there any bugs on scCombination or it is because of the Seurat version? Or I need to add some extra arguments? Thank you.