search

satijalab / seurat-object

https://satijalab.github.io/seurat-object/
Other
24 stars 25 forks source link

Error converting Assay5 to Assay #206

Closed alexvpickering closed 5 months ago

alexvpickering commented 6 months ago

Following the vignette, I keep getting error Error in .subset2(x, i, exact = exact) : subscript out of bounds:

library(Seurat)
library(SeuratData)
options(Seurat.object.assay.version = "v5")
pbmc3k <- LoadData("pbmc3k")

class(pbmc3k[["RNA"]])
# [1] "Assay5"
# attr(,"package")
# [1] "SeuratObject"
as(object = pbmc3k[["RNA"]], Class = "Assay")

Warning: No layers found matching search pattern provided
Warning: No layers found matching search pattern provided
Error in .subset2(x, i, exact = exact) : subscript out of bounds
In addition: Warning message:
Layer ‘scale.data’ is empty

Have tried with a few different datasets and getting the same error.

sessionInfo()
R version 4.4.0 (2024-04-24)
Platform: x86_64-pc-linux-gnu
Running under: Ubuntu 22.04.1 LTS

Matrix products: default
BLAS:   /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3 
LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.20.so;  LAPACK version 3.10.0

locale:
 [1] LC_CTYPE=en_CA.UTF-8       LC_NUMERIC=C               LC_TIME=en_CA.UTF-8        LC_COLLATE=en_CA.UTF-8     LC_MONETARY=en_CA.UTF-8   
 [6] LC_MESSAGES=en_CA.UTF-8    LC_PAPER=en_CA.UTF-8       LC_NAME=C                  LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_CA.UTF-8 LC_IDENTIFICATION=C       

time zone: America/Vancouver
tzcode source: system (glibc)

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

other attached packages:
[1] pbmc3k.SeuratData_3.1.4 SeuratData_0.2.2.9001   Seurat_5.1.0            SeuratObject_5.0.2      sp_2.1-4               

loaded via a namespace (and not attached):
  [1] deldir_2.0-4           pbapply_1.7-2          gridExtra_2.3          remotes_2.5.0          rlang_1.1.3            magrittr_2.0.3        
  [7] RcppAnnoy_0.0.22       matrixStats_1.3.0      ggridges_0.5.6         compiler_4.4.0         spatstat.geom_3.2-9    png_0.1-8             
 [13] vctrs_0.6.5            reshape2_1.4.4         stringr_1.5.1          crayon_1.5.2           pkgconfig_2.0.3        fastmap_1.2.0         
 [19] utf8_1.2.4             promises_1.3.0         purrr_1.0.2            jsonlite_1.8.8         goftest_1.2-3          later_1.3.2           
 [25] spatstat.utils_3.0-4   irlba_2.3.5.1          parallel_4.4.0         cluster_2.1.6          R6_2.5.1               ica_1.0-3             
 [31] stringi_1.8.4          RColorBrewer_1.1-3     spatstat.data_3.0-4    reticulate_1.36.1      parallelly_1.37.1      lmtest_0.9-40         
 [37] scattermore_1.2        Rcpp_1.0.12            tensor_1.5             future.apply_1.11.2    zoo_1.8-12             sctransform_0.4.1     
 [43] httpuv_1.6.15          Matrix_1.6-5           splines_4.4.0          igraph_2.0.3           tidyselect_1.2.1       abind_1.4-5           
 [49] spatstat.random_3.2-3  codetools_0.2-19       miniUI_0.1.1.1         spatstat.explore_3.2-7 curl_5.2.1             listenv_0.9.1         
 [55] lattice_0.22-5         tibble_3.2.1           plyr_1.8.9             shiny_1.8.1.1          ROCR_1.0-11            Rtsne_0.17            
 [61] future_1.33.2          fastDummies_1.7.3      survival_3.5-8         polyclip_1.10-6        fitdistrplus_1.1-11    pillar_1.9.0          
 [67] KernSmooth_2.23-22     plotly_4.10.4          generics_0.1.3         RcppHNSW_0.6.0         ggplot2_3.5.1          munsell_0.5.1         
 [73] scales_1.3.0           globals_0.16.3         xtable_1.8-4           glue_1.7.0             lazyeval_0.2.2         tools_4.4.0           
 [79] data.table_1.15.4      RSpectra_0.16-1        RANN_2.6.1             leiden_0.4.3.1         dotCall64_1.1-1        cowplot_1.1.3         
 [85] grid_4.4.0             tidyr_1.3.1            colorspace_2.1-0       nlme_3.1-163           patchwork_1.2.0        cli_3.6.2             
 [91] rappdirs_0.3.3         spatstat.sparse_3.0-3  spam_2.10-0            fansi_1.0.6            viridisLite_0.4.2      dplyr_1.1.4           
 [97] uwot_0.2.2             gtable_0.3.5           digest_0.6.35          progressr_0.14.0       ggrepel_0.9.5          htmlwidgets_1.6.4     
[103] htmltools_0.5.8.1      lifecycle_1.0.4        httr_1.4.7             mime_0.12              MASS_7.3-60           
>
dcollins15 commented 6 months ago

Thanks for reaching out @alexvpickering

It's surprising that we'd be doing this in a vignette, AFAIK this is not a supported operation. What vignette did you get this from?

alexvpickering commented 6 months ago

Thanks for reaching out @alexvpickering

It's surprising that we'd be doing this in a vignette, AFAIK this is not a supported operation. What vignette did you get this from?

https://satijalab.org/seurat/articles/seurat5_essential_commands

luzzhou commented 6 months ago

Following the vignette, I keep getting error Error in .subset2(x, i, exact = exact) : subscript out of bounds:

library(Seurat)
library(SeuratData)
options(Seurat.object.assay.version = "v5")
pbmc3k <- LoadData("pbmc3k")

class(pbmc3k[["RNA"]])
# [1] "Assay5"
# attr(,"package")
# [1] "SeuratObject"
as(object = pbmc3k[["RNA"]], Class = "Assay")

Warning: No layers found matching search pattern provided
Warning: No layers found matching search pattern provided
Error in .subset2(x, i, exact = exact) : subscript out of bounds
In addition: Warning message:
Layer ‘scale.data’ is empty

Have tried with a few different datasets and getting the same error.

sessionInfo()
R version 4.4.0 (2024-04-24)
Platform: x86_64-pc-linux-gnu
Running under: Ubuntu 22.04.1 LTS

Matrix products: default
BLAS:   /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3 
LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.20.so;  LAPACK version 3.10.0

locale:
 [1] LC_CTYPE=en_CA.UTF-8       LC_NUMERIC=C               LC_TIME=en_CA.UTF-8        LC_COLLATE=en_CA.UTF-8     LC_MONETARY=en_CA.UTF-8   
 [6] LC_MESSAGES=en_CA.UTF-8    LC_PAPER=en_CA.UTF-8       LC_NAME=C                  LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_CA.UTF-8 LC_IDENTIFICATION=C       

time zone: America/Vancouver
tzcode source: system (glibc)

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

other attached packages:
[1] pbmc3k.SeuratData_3.1.4 SeuratData_0.2.2.9001   Seurat_5.1.0            SeuratObject_5.0.2      sp_2.1-4               

loaded via a namespace (and not attached):
  [1] deldir_2.0-4           pbapply_1.7-2          gridExtra_2.3          remotes_2.5.0          rlang_1.1.3            magrittr_2.0.3        
  [7] RcppAnnoy_0.0.22       matrixStats_1.3.0      ggridges_0.5.6         compiler_4.4.0         spatstat.geom_3.2-9    png_0.1-8             
 [13] vctrs_0.6.5            reshape2_1.4.4         stringr_1.5.1          crayon_1.5.2           pkgconfig_2.0.3        fastmap_1.2.0         
 [19] utf8_1.2.4             promises_1.3.0         purrr_1.0.2            jsonlite_1.8.8         goftest_1.2-3          later_1.3.2           
 [25] spatstat.utils_3.0-4   irlba_2.3.5.1          parallel_4.4.0         cluster_2.1.6          R6_2.5.1               ica_1.0-3             
 [31] stringi_1.8.4          RColorBrewer_1.1-3     spatstat.data_3.0-4    reticulate_1.36.1      parallelly_1.37.1      lmtest_0.9-40         
 [37] scattermore_1.2        Rcpp_1.0.12            tensor_1.5             future.apply_1.11.2    zoo_1.8-12             sctransform_0.4.1     
 [43] httpuv_1.6.15          Matrix_1.6-5           splines_4.4.0          igraph_2.0.3           tidyselect_1.2.1       abind_1.4-5           
 [49] spatstat.random_3.2-3  codetools_0.2-19       miniUI_0.1.1.1         spatstat.explore_3.2-7 curl_5.2.1             listenv_0.9.1         
 [55] lattice_0.22-5         tibble_3.2.1           plyr_1.8.9             shiny_1.8.1.1          ROCR_1.0-11            Rtsne_0.17            
 [61] future_1.33.2          fastDummies_1.7.3      survival_3.5-8         polyclip_1.10-6        fitdistrplus_1.1-11    pillar_1.9.0          
 [67] KernSmooth_2.23-22     plotly_4.10.4          generics_0.1.3         RcppHNSW_0.6.0         ggplot2_3.5.1          munsell_0.5.1         
 [73] scales_1.3.0           globals_0.16.3         xtable_1.8-4           glue_1.7.0             lazyeval_0.2.2         tools_4.4.0           
 [79] data.table_1.15.4      RSpectra_0.16-1        RANN_2.6.1             leiden_0.4.3.1         dotCall64_1.1-1        cowplot_1.1.3         
 [85] grid_4.4.0             tidyr_1.3.1            colorspace_2.1-0       nlme_3.1-163           patchwork_1.2.0        cli_3.6.2             
 [91] rappdirs_0.3.3         spatstat.sparse_3.0-3  spam_2.10-0            fansi_1.0.6            viridisLite_0.4.2      dplyr_1.1.4           
 [97] uwot_0.2.2             gtable_0.3.5           digest_0.6.35          progressr_0.14.0       ggrepel_0.9.5          htmlwidgets_1.6.4     
[103] htmltools_0.5.8.1      lifecycle_1.0.4        httr_1.4.7             mime_0.12              MASS_7.3-60           
>

Hi, @alexvpickering. I have the same problem now, have you solved it?

dcollins15 commented 6 months ago

Apologies for the delay, and thank you for the information 😄

I'd strongly recommend that you use CreateAssayObject to instantiate a new v3 assay instead of converting with as:

assay_v3 <- CreateAssayObject(
  counts = pbmc3k[["RNA"]]$counts,
  # optional
  data = pbmc3k[["RNA"]]$data
)

pbmc3k[["RNA"]] <- assay_v3

The issue you've both encountered can be resolved by calling ScaleData on pbmc3k, but this also highlights why you should avoid using as for this conversion. The Assay and Assay5 classes are only isomorphic if the latter has layers corresponding to the three slots defined in the former (counts, data, scale.data). This alignment is ensured when using Seurat’s standard preprocessing workflows, but it might not hold if your assay has non-standard layer names or is missing expected layers.

Using CreateAssayObject ensures that your assay object is correctly instantiated with the expected v3 structure. We should have an update for our vignettes soon that will provide better guidance 👍

alexvpickering commented 6 months ago

Thank you @dcollins15 - I ended up taking the more involved but ultimately more durable approach and switched over our app to be compatible with v3 and v5 Assays. I noticed a couple oddities that I'll post as separate issues in case they are unintentional.

samuel-marsh commented 5 months ago

Hi @dcollins15

Just to chime in on something that would be helpful if possible in implementing a new way of doing this (or updating the instructions at least) is that if one does the conversion/replacement using CreateAssayObject then the assay meta data (variable features etc) are lost. If there is way to add feature to CreateAssayObject/CreateAssay5Object to transfer assay meta data that would be great!

Best, Sam

dcollins15 commented 5 months ago

Great call @samuel-marsh - I'll add this to our roadmap 👍

samuel-marsh commented 5 months ago

Great!👍

mojaveazure commented 5 months ago

v5 assays were designed to be behaviorly similar to and forwards-compatible with v3 assays, but are fundamentally different and violate many of the conditions of a v3 Assay. These conditions include

While direct conversions from a v5 assay to a v3 assay may be possible under "typical" circumstances, it's not trivial to check and verify that a v5 assays is able to be seamlessly converted to a v3 assay. SeuratObject has an internal converter used in very specific and controlled cases, and that converter should not have been exposed (our bad)

@dcollins15 during the v5 development cycle, I made a number of improvements to handling meta data additions at the Seurat level; we should work on bringing these down to the assay level so that users can easily transfer feature-level meta data between assays that have some overlap in features. However, given the differences between v3 and v5 assays and explicit lack of backwards compatibility from v5 to v3, a direct v5 to v3 converter is out-of-scope and should not be pursued

samuel-marsh commented 5 months ago

@mojaveazure that all totally makes sense, thanks for the details!!

HumaNaz0011 commented 5 months ago

Hi team! I mistakenly updated seurat to V5. after that i was having a lot of problems in analyzing my data. So I switched back to V4. Now i am not able to process my object that has assay5. when i tried to plot voilin plot or other downstrem analysis i got " Error: 'RNA' is not an assay" . although it is present in object. as "S4 object of class assay5". I tried different ways mentioned above to convert Assay5 to Assay but not successful(may be due to current seurat V ‘4.4.0’.) how to solve this problem. Is any way to convert "Assay5" to "Assay" using seurat V ‘4.4.0’? thanks in advance