errors integrating multiple 10X Multiome datasets

[mzz92]

I have three 10x Multiome datasets that I would like to do an integrated analysis on but am getting various errors with FindIntegrationAnchors(), such as "The default assay is slated to be removed, please change the default assay". Looking through past posts such as signac posts 143, 181, 182, I've tried various corrections but have failed to resolve this.

First, I followed the "vignettes/weighted_nearest_neighbor_analysis.Rmd" to produce 3 WNN 10x Multiome Seurat objects and saved them to .rds files. Then I adapted the "vignettes/integrate_atac.Rmd" workflow in this way:

library(Signac)
library(Seurat)
library(ggplot2)

pbmc.W1 <- readRDS("W1_multiome.rds")
pbmc.W2 <- readRDS("W2_multiome.rds")
pbmc.W3 <- readRDS("W3_multiome.rds")

pbmc.W1$dataset <- "W1"
pbmc.W2$dataset <- "W2"
pbmc.W3$dataset <- "W3"

pbmc.combined <- merge(x=pbmc.W1, y=c(pbmc.W2,pbmc.W3) )

pbmc.combined <- FindTopFeatures(pbmc.combined, min.cutoff = 10)
pbmc.combined <- RunTFIDF(pbmc.combined)
pbmc.combined <- RunSVD(pbmc.combined)
pbmc.combined <- RunUMAP(pbmc.combined, reduction = "lsi", dims = 2:30)

integration.anchors <- FindIntegrationAnchors(
  object.list = list(pbmc.W2,pbmc.W2,pbmc.W3),
  anchor.features = rownames(pbmc.combined),
  reduction = "rlsi",
  dims = 2:30
)

At this point, I get this error message:

Computing within dataset neighborhoods
  |++++++++++++++++++++++++++++++++++++++++++++++++++| 100% elapsed=01m 29s
Finding all pairwise anchors
  |                                                  | 0 % ~calculating  Error in DietSeurat(object = object.list[[i]], assays = assay[i], features = anchor.features,  : 
  The default assay is slated to be removed, please change the default assay
In addition: Warning message:
In CheckDuplicateCellNames(object.list = object.list) :
  Some cell names are duplicated across objects provided. Renaming to enforce unique cell names.

These are the contents of pbmc.W1 (pbmc.W2 and pbmc.W3 are similar) and pbmc.combined:

> pbmc.W1
An object of class Seurat 
374544 features across 6791 samples within 4 assays 
Active assay: RNA (36601 features, 0 variable features)
 3 other assays present: ATAC, peaks, SCT
 2 dimensional reductions calculated: pca, lsi

> pbmc.combined
An object of class Seurat 
413308 features across 31687 samples within 4 assays 
Active assay: RNA (36601 features, 0 variable features)
 3 other assays present: ATAC, peaks, SCT
 2 dimensional reductions calculated: lsi, umap

My session info:

> sessionInfo()
R version 4.0.3 (2020-10-10)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: CentOS Linux 7 (Core)

Matrix products: default
BLAS/LAPACK: /usr/lib64/libopenblas-r0.3.3.so

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8    
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
 [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] ggplot2_3.3.5      SeuratObject_4.0.2 Seurat_4.0.4       Signac_1.4.0      

loaded via a namespace (and not attached):
  [1] Rtsne_0.15             colorspace_2.0-2       deldir_0.2-10         
  [4] ellipsis_0.3.2         ggridges_0.5.3         lsa_0.73.2            
  [7] XVector_0.30.0         GenomicRanges_1.42.0   rstudioapi_0.13       
 [10] spatstat.data_2.1-0    farver_2.1.0           leiden_0.3.9          
 [13] listenv_0.8.0          SnowballC_0.7.0        ggrepel_0.9.1         
 [16] RSpectra_0.16-0        fansi_0.5.0            codetools_0.2-18      
 [19] splines_4.0.3          docopt_0.7.1           RcppRoll_0.3.0        
 [22] polyclip_1.10-0        jsonlite_1.7.2         Rsamtools_2.6.0       
 [25] ica_1.0-2              cluster_2.1.2          png_0.1-7             
 [28] uwot_0.1.10            ggforce_0.3.3          shiny_1.7.0           
 [31] sctransform_0.3.2      spatstat.sparse_2.0-0  compiler_4.0.3        
 [34] httr_1.4.2             assertthat_0.2.1       Matrix_1.3-4          
 [37] fastmap_1.1.0          lazyeval_0.2.2         tweenr_1.0.2          
 [40] later_1.3.0            htmltools_0.5.2        tools_4.0.3           
 [43] igraph_1.2.6           gtable_0.3.0           glue_1.4.2            
 [46] GenomeInfoDbData_1.2.4 RANN_2.6.1             reshape2_1.4.4        
 [49] dplyr_1.0.7            fastmatch_1.1-3        Rcpp_1.0.7            
 [52] slam_0.1-48            scattermore_0.7        vctrs_0.3.8           
 [55] Biostrings_2.58.0      nlme_3.1-153           ggseqlogo_0.1         
 [58] lmtest_0.9-38          stringr_1.4.0          globals_0.14.0        
 [61] mime_0.11              miniUI_0.1.1.1         lifecycle_1.0.0       
 [64] irlba_2.3.3            goftest_1.2-2          future_1.22.1         
 [67] zlibbioc_1.36.0        MASS_7.3-54            zoo_1.8-9             
 [70] scales_1.1.1           spatstat.core_2.3-0    promises_1.2.0.1      
 [73] spatstat.utils_2.2-0   parallel_4.0.3         RColorBrewer_1.1-2    
 [76] reticulate_1.22        pbapply_1.5-0          gridExtra_2.3         
 [79] rpart_4.1-15           stringi_1.7.4          S4Vectors_0.28.1      
 [82] BiocGenerics_0.36.1    BiocParallel_1.24.1    GenomeInfoDb_1.26.7   
 [85] rlang_0.4.11           pkgconfig_2.0.3        matrixStats_0.61.0    
 [88] bitops_1.0-7           qlcMatrix_0.9.7        lattice_0.20-45       
 [91] ROCR_1.0-11            purrr_0.3.4            tensor_1.5            
 [94] patchwork_1.1.1        htmlwidgets_1.5.4      cowplot_1.1.1         
 [97] tidyselect_1.1.1       parallelly_1.28.1      RcppAnnoy_0.0.19      
[100] plyr_1.8.6             magrittr_2.0.1         R6_2.5.1              
[103] IRanges_2.24.1         generics_0.1.0         DBI_1.1.1             
[106] withr_2.4.2            mgcv_1.8-37            pillar_1.6.2          
[109] fitdistrplus_1.1-5     survival_3.2-13        abind_1.4-5           
[112] RCurl_1.98-1.5         tibble_3.1.4           future.apply_1.8.1    
[115] crayon_1.4.1           KernSmooth_2.23-20     utf8_1.2.2            
[118] spatstat.geom_2.2-2    plotly_4.9.4.1         grid_4.0.3            
[121] data.table_1.14.0      sparsesvd_0.2          digest_0.6.28         
[124] xtable_1.8-4           tidyr_1.1.3            httpuv_1.6.3          
[127] stats4_4.0.3           munsell_0.5.0          viridisLite_0.4.0     

If I change the DefaultAssay, I get a different error message:

> DefaultAssay(pbmc.combined) <- "RNA"
> DefaultAssay(pbmc.W1) <- "RNA"
> DefaultAssay(pbmc.W2) <- "RNA"
> DefaultAssay(pbmc.W3) <- "RNA"
> 
> integration.anchors <- FindIntegrationAnchors(
+   object.list = list(pbmc.W1,pbmc.W2,pbmc.W3),
+   anchor.features = rownames(pbmc.combined),
+   reduction = "rlsi",
+   dims = 2:30
+ )
Computing within dataset neighborhoods
  |++++++++++++++++++++++++++++++++++++++++++++++++++| 100% elapsed=01m 30s
Finding all pairwise anchors
  |                                                  | 0 % ~calculating  Error: Cannot find 'lsi' in this Seurat object
In addition: Warning message:
In CheckDuplicateCellNames(object.list = object.list) :
  Some cell names are duplicated across objects provided. Renaming to enforce unique cell names.

Looking at previous posts, I've even tried to reformat chrXXX:XXX-XXX to no avail:

> rownames(pbmc.W1@assays$peaks@counts) <- sub("-",":",rownames(pbmc.W1@assays$peaks))
> rownames(pbmc.W1@assays$peaks@data) <- sub("-",":",rownames(pbmc.W1@assays$peaks))
> 
> rownames(pbmc.W2@assays$peaks@counts) <- sub("-",":",rownames(pbmc.W2@assays$peaks))
> rownames(pbmc.W2@assays$peaks@data) <- sub("-",":",rownames(pbmc.W2@assays$peaks))
> 
> rownames(pbmc.W3@assays$peaks@counts) <- sub("-",":",rownames(pbmc.W3@assays$peaks))
> rownames(pbmc.W3@assays$peaks@data) <- sub("-",":",rownames(pbmc.W3@assays$peaks))
> 
> rownames(pbmc.combined@assays$peaks@counts) <- sub("-",":",rownames(pbmc.combined@assays$peaks))
> rownames(pbmc.combined@assays$peaks@data) <- sub("-",":",rownames(pbmc.combined@assays$peaks))
> integration.anchors <- FindIntegrationAnchors(
+   object.list = list(pbmc.W1,pbmc.W2,pbmc.W3),
+   anchor.features = rownames(pbmc.combined),
+   reduction = "rlsi",
+   dims = 2:30
+ )
Computing within dataset neighborhoods
  |++++++++++++++++++++++++++++++++++++++++++++++++++| 100% elapsed=01m 28s
Finding all pairwise anchors
  |                                                  | 0 % ~calculating  Error in DietSeurat(object = object.list[[i]], assays = assay[i], features = anchor.features,  : 
  The default assay is slated to be removed, please change the default assay
In addition: Warning message:
In CheckDuplicateCellNames(object.list = object.list) :
  Some cell names are duplicated across objects provided. Renaming to enforce unique cell names.

Any pointers would be much appreciated!