bug: MAST returns error after Seurat SCTransform, v2 regularization #175

grhogg · 2022-09-30T17:45:28Z

Hello,

Thank you for your amazing package!

I have been running into an error when trying to run FindMarkers on a Seurat Object with the MAST algorithm, using their newer v2 regularization pipeline. Any help would be greatly appreciated.

error is generated after calling MAST in findmarkers. Other tests like Wilcox still work fine

Error in .nextMethod(.Object = .Object, ... = ..., design = design) :
invalid name for slot of class “BayesGLMlike”: norm.method

R base 4.2.1
Seurat version 4.2
Mast version 1.22
SCTransform version 0.3.5

Example with Seurat Dataset:


library(Seurat)
library(SeuratData)
library(patchwork)
library(dplyr)
library(ggplot2)
library(MAST)

InstallData("ifnb")

# load dataset

LoadData("ifnb")

# split the dataset into a list of two seurat objects (stim and CTRL)

ifnb.list <- SplitObject(ifnb, split.by = "stim")

ctrl <- ifnb.list[["CTRL"]]
stim <- ifnb.list[["STIM"]]

ctrl <- SCTransform(ctrl, vst.flavor = "v2")
stim <- SCTransform(stim, vst.flavor = "v2") 

ifnb.list <- list(ctrl = ctrl, stim = stim)

#integrate datasets

features <- SelectIntegrationFeatures(object.list = ifnb.list, nfeatures = 3000)
ifnb.list <- PrepSCTIntegration(object.list = ifnb.list, anchor.features = features)
immune.anchors <- FindIntegrationAnchors(object.list = ifnb.list, normalization.method = "SCT",
                                         anchor.features = features)
immune.combined.sct <- IntegrateData(anchorset = immune.anchors, normalization.method = "SCT")


#assign stim conditions

immune.combined.sct$celltype.stim <- paste(immune.combined.sct$seurat_annotations, immune.combined.sct$stim,
                                           sep = "_")
Idents(immune.combined.sct) <- "celltype.stim"

#correct SCT Counts

immune.combined.sct <- PrepSCTFindMarkers(immune.combined.sct)

#find Markers with MAST algorithm

b.interferon.response <- FindMarkers(immune.combined.sct, assay = "SCT", ident.1 = "B_STIM", ident.2 = "B_CTRL",
                                      test.use = "MAST")

Error in .nextMethod(.Object = .Object, ... = ..., design = design) :
invalid name for slot of class “BayesGLMlike”: norm.method

sessionInfo()

_R version 4.2.1 (2022-06-23 ucrt)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 22000)

Matrix products: default

locale:
[1] LC_COLLATE=English_United States.utf8 LC_CTYPE=English_United States.utf8
[3] LC_MONETARY=English_United States.utf8 LC_NUMERIC=C
[5] LC_TIME=English_United States.utf8

attached base packages:
[1] stats4 stats graphics grDevices utils datasets methods base

other attached packages:
[1] ifnb.SeuratData_3.1.0 SeuratData_0.2.2 MAST_1.22.0
[4] SingleCellExperiment_1.18.0 SummarizedExperiment_1.26.1 Biobase_2.56.0
[7] GenomicRanges_1.48.0 GenomeInfoDb_1.32.4 IRanges_2.30.1
[10] S4Vectors_0.34.0 BiocGenerics_0.42.0 MatrixGenerics_1.8.1
[13] matrixStats_0.62.0 ggplot2_3.3.6 dplyr_1.0.10
[16] patchwork_1.1.2 sp_1.5-0 SeuratObject_4.1.2
[19] Seurat_4.2.0

loaded via a namespace (and not attached):
[1] plyr_1.8.7 igraph_1.3.5 lazyeval_0.2.2 splines_4.2.1
[5] listenv_0.8.0 scattermore_0.8 usethis_2.1.6 digest_0.6.29
[9] htmltools_0.5.3 fansi_1.0.3 magrittr_2.0.3 memoise_2.0.1
[13] tensor_1.5 cluster_2.1.3 ROCR_1.0-11 remotes_2.4.2
[17] globals_0.16.1 spatstat.sparse_2.1-1 prettyunits_1.1.1 colorspace_2.0-3
[21] rappdirs_0.3.3 ggrepel_0.9.1 xfun_0.33 RCurl_1.98-1.8
[25] callr_3.7.2 crayon_1.5.2 jsonlite_1.8.0 progressr_0.11.0
[29] spatstat.data_2.2-0 survival_3.3-1 zoo_1.8-11 glue_1.6.2
[33] polyclip_1.10-0 gtable_0.3.1 zlibbioc_1.42.0 XVector_0.36.0
[37] leiden_0.4.3 DelayedArray_0.22.0 pkgbuild_1.3.1 future.apply_1.9.1
[41] abind_1.4-5 scales_1.2.1 DBI_1.1.3 spatstat.random_2.2-0
[45] miniUI_0.1.1.1 Rcpp_1.0.9 viridisLite_0.4.1 xtable_1.8-4
[49] reticulate_1.26 spatstat.core_2.4-4 profvis_0.3.7 htmlwidgets_1.5.4
[53] httr_1.4.4 RColorBrewer_1.1-3 ellipsis_0.3.2 ica_1.0-3
[57] urlchecker_1.0.1 pkgconfig_2.0.3 uwot_0.1.14 deldir_1.0-6
[61] utf8_1.2.2 tidyselect_1.1.2 rlang_1.0.6 reshape2_1.4.4
[65] later_1.3.0 munsell_0.5.0 tools_4.2.1 cachem_1.0.6
[69] cli_3.4.1 generics_0.1.3 devtools_2.4.4 ggridges_0.5.4
[73] evaluate_0.16 stringr_1.4.1 fastmap_1.1.0 yaml_2.3.5
[77] goftest_1.2-3 processx_3.7.0 knitr_1.40 fs_1.5.2
[81] fitdistrplus_1.1-8 purrr_0.3.4 RANN_2.6.1 sparseMatrixStats_1.8.0
[85] pbapply_1.5-0 future_1.28.0 nlme_3.1-157 mime_0.12
[89] compiler_4.2.1 rstudioapi_0.14 plotly_4.10.0 curl_4.3.2
[93] png_0.1-7 spatstat.utils_2.3-1 tibble_3.1.8 glmGamPoi_1.8.0
[97] stringi_1.7.8 ps_1.7.1 rgeos_0.5-9 lattice_0.20-45
[101] Matrix_1.5-1 vctrs_0.4.2 pillar_1.8.1 lifecycle_1.0.2
[105] spatstat.geom_2.4-0 lmtest_0.9-40 RcppAnnoy_0.0.19 bitops_1.0-7
[109] data.table_1.14.2 cowplot_1.1.1 irlba_2.3.5 httpuv_1.6.6
[113] R6_2.5.1 promises_1.2.0.1 KernSmooth_2.23-20 gridExtra_2.3
[117] parallelly_1.32.1 sessioninfo_1.2.2 codetools_0.2-18 MASS_7.3-57
[121] assertthat_0.2.1 pkgload_1.3.0 rprojroot_2.0.3 withr_2.5.0
[125] sctransform_0.3.5 GenomeInfoDbData_1.2.8 mgcv_1.8-40 parallel_4.2.1
[129] grid_4.2.1 rpart_4.1.16 tidyr_1.2.1 DelayedMatrixStats_1.18.1
[133] rmarkdown_2.16 Rtsne_0.16 shiny_1.7.2_

The text was updated successfully, but these errors were encountered:

amcdavid · 2022-09-30T19:05:38Z

There's a typo or invalid argument being passed to MAST from `FindAllMarkers`, a bug in Seurat. Might get a hint about where this is happening by looking at the `traceback()`.

…

On Fri, Sep 30, 2022 at 10:45 AM grhogg ***@***.***> wrote: Hello, Thank you for your amazing package! I have been running into an error when trying to run FindMarkers on a Seurat Object with the MAST algorithm, using their newer v2 regularization pipeline. Any help would be greatly appreciated. error is generated after calling MAST in findmarkers. Other tests like Wilcox still work fine Error in .nextMethod(.Object = .Object, ... = ..., design = design) : invalid name for slot of class “BayesGLMlike”: norm.method R base 4.2.1 Seurat version 4.2 Mast version 1.22 SCTransform version 0.3.5 Example with Seurat Dataset: library(Seurat) library(SeuratData) library(patchwork) library(dplyr) library(ggplot2) library(MAST) InstallData("ifnb") # load dataset LoadData("ifnb") # split the dataset into a list of two seurat objects (stim and CTRL) ifnb.list <- SplitObject(ifnb, split.by = "stim") ctrl <- ifnb.list[["CTRL"]] stim <- ifnb.list[["STIM"]] ctrl <- SCTransform(ctrl, vst.flavor = "v2") stim <- SCTransform(stim, vst.flavor = "v2") ifnb.list <- list(ctrl = ctrl, stim = stim) #integrate datasets features <- SelectIntegrationFeatures(object.list = ifnb.list, nfeatures = 3000) ifnb.list <- PrepSCTIntegration(object.list = ifnb.list, anchor.features = features) immune.anchors <- FindIntegrationAnchors(object.list = ifnb.list, normalization.method = "SCT", anchor.features = features) immune.combined.sct <- IntegrateData(anchorset = immune.anchors, normalization.method = "SCT") #assign stim conditions immune.combined.sct$celltype.stim <- paste(immune.combined.sct$seurat_annotations, immune.combined.sct$stim, sep = "_") Idents(immune.combined.sct) <- "celltype.stim" #correct SCT Counts immune.combined.sct <- PrepSCTFindMarkers(immune.combined.sct) #find Markers with MAST algorithm b.interferon.response <- FindMarkers(immune.combined.sct, assay = "SCT", ident.1 = "B_STIM", ident.2 = "B_CTRL", test.use = "MAST") Error in .nextMethod(.Object = .Object, ... = ..., design = design) : invalid name for slot of class “BayesGLMlike”: norm.method sessionInfo() _R version 4.2.1 (2022-06-23 ucrt) Platform: x86_64-w64-mingw32/x64 (64-bit) Running under: Windows 10 x64 (build 22000) Matrix products: default locale: [1] LC_COLLATE=English_United States.utf8 LC_CTYPE=English_United States.utf8 [3] LC_MONETARY=English_United States.utf8 LC_NUMERIC=C [5] LC_TIME=English_United States.utf8 attached base packages: [1] stats4 stats graphics grDevices utils datasets methods base other attached packages: [1] ifnb.SeuratData_3.1.0 SeuratData_0.2.2 MAST_1.22.0 [4] SingleCellExperiment_1.18.0 SummarizedExperiment_1.26.1 Biobase_2.56.0 [7] GenomicRanges_1.48.0 GenomeInfoDb_1.32.4 IRanges_2.30.1 [10] S4Vectors_0.34.0 BiocGenerics_0.42.0 MatrixGenerics_1.8.1 [13] matrixStats_0.62.0 ggplot2_3.3.6 dplyr_1.0.10 [16] patchwork_1.1.2 sp_1.5-0 SeuratObject_4.1.2 [19] Seurat_4.2.0 loaded via a namespace (and not attached): [1] plyr_1.8.7 igraph_1.3.5 lazyeval_0.2.2 splines_4.2.1 [5] listenv_0.8.0 scattermore_0.8 usethis_2.1.6 digest_0.6.29 [9] htmltools_0.5.3 fansi_1.0.3 magrittr_2.0.3 memoise_2.0.1 [13] tensor_1.5 cluster_2.1.3 ROCR_1.0-11 remotes_2.4.2 [17] globals_0.16.1 spatstat.sparse_2.1-1 prettyunits_1.1.1 colorspace_2.0-3 [21] rappdirs_0.3.3 ggrepel_0.9.1 xfun_0.33 RCurl_1.98-1.8 [25] callr_3.7.2 crayon_1.5.2 jsonlite_1.8.0 progressr_0.11.0 [29] spatstat.data_2.2-0 survival_3.3-1 zoo_1.8-11 glue_1.6.2 [33] polyclip_1.10-0 gtable_0.3.1 zlibbioc_1.42.0 XVector_0.36.0 [37] leiden_0.4.3 DelayedArray_0.22.0 pkgbuild_1.3.1 future.apply_1.9.1 [41] abind_1.4-5 scales_1.2.1 DBI_1.1.3 spatstat.random_2.2-0 [45] miniUI_0.1.1.1 Rcpp_1.0.9 viridisLite_0.4.1 xtable_1.8-4 [49] reticulate_1.26 spatstat.core_2.4-4 profvis_0.3.7 htmlwidgets_1.5.4 [53] httr_1.4.4 RColorBrewer_1.1-3 ellipsis_0.3.2 ica_1.0-3 [57] urlchecker_1.0.1 pkgconfig_2.0.3 uwot_0.1.14 deldir_1.0-6 [61] utf8_1.2.2 tidyselect_1.1.2 rlang_1.0.6 reshape2_1.4.4 [65] later_1.3.0 munsell_0.5.0 tools_4.2.1 cachem_1.0.6 [69] cli_3.4.1 generics_0.1.3 devtools_2.4.4 ggridges_0.5.4 [73] evaluate_0.16 stringr_1.4.1 fastmap_1.1.0 yaml_2.3.5 [77] goftest_1.2-3 processx_3.7.0 knitr_1.40 fs_1.5.2 [81] fitdistrplus_1.1-8 purrr_0.3.4 RANN_2.6.1 sparseMatrixStats_1.8.0 [85] pbapply_1.5-0 future_1.28.0 nlme_3.1-157 mime_0.12 [89] compiler_4.2.1 rstudioapi_0.14 plotly_4.10.0 curl_4.3.2 [93] png_0.1-7 spatstat.utils_2.3-1 tibble_3.1.8 glmGamPoi_1.8.0 [97] stringi_1.7.8 ps_1.7.1 rgeos_0.5-9 lattice_0.20-45 [101] Matrix_1.5-1 vctrs_0.4.2 pillar_1.8.1 lifecycle_1.0.2 [105] spatstat.geom_2.4-0 lmtest_0.9-40 RcppAnnoy_0.0.19 bitops_1.0-7 [109] data.table_1.14.2 cowplot_1.1.1 irlba_2.3.5 httpuv_1.6.6 [113] R6_2.5.1 promises_1.2.0.1 KernSmooth_2.23-20 gridExtra_2.3 [117] parallelly_1.32.1 sessioninfo_1.2.2 codetools_0.2-18 MASS_7.3-57 [121] assertthat_0.2.1 pkgload_1.3.0 rprojroot_2.0.3 withr_2.5.0 [125] sctransform_0.3.5 GenomeInfoDbData_1.2.8 mgcv_1.8-40 parallel_4.2.1 [129] grid_4.2.1 rpart_4.1.16 tidyr_1.2.1 DelayedMatrixStats_1.18.1 [133] rmarkdown_2.16 Rtsne_0.16 shiny_1.7.2_ — Reply to this email directly, view it on GitHub <#175>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AALLAHUF3PUWUUNKDQ6HP73WA4RMJANCNFSM6AAAAAAQZ7EMLE> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

grhogg · 2022-09-30T19:49:05Z

Thanks for your quick response! I have posted traceback() output below. Would this bug be better reported on the Seurat Github? Thanks again.

traceback()

18: stop(sprintf(ngettext(sum(is.na(which)), "invalid name for slot of class %s: %s",
"invalid names for slots of class %s: %s"), dQuote(Class),
paste(snames[is.na(which)], collapse = ", ")), domain = NA)

17: .nextMethod(.Object = .Object, ... = ..., design = design)

16: callNextMethod()

15: .local(.Object, ...)

14: .nextMethod(.Object = .Object, ... = ...)

13: callNextMethod()

12: initialize(value, ...)

11: initialize(value, ...)

10: new(method, design = colData(sca), formula = Formula, priorVar = priorVar,
priorDOF = priorDOF, ...)

9: MAST::zlm(formula = fmla, sca = sca, ...)

8: MASTDETest(data.use = data.use, cells.1 = cells.1, cells.2 = cells.2,
latent.vars = latent.vars, verbose = verbose, ...)

7: PerformDE(object = object, cells.1 = cells.1, cells.2 = cells.2,
features = features, test.use = test.use, verbose = verbose,
min.cells.feature = min.cells.feature, latent.vars = latent.vars,
densify = densify, ...)

6: FindMarkers.default(object = data.use, slot = data.slot, counts = counts,
cells.1 = cells.1, cells.2 = cells.2, features = features,
logfc.threshold = logfc.threshold, test.use = test.use, min.pct = min.pct,
min.diff.pct = min.diff.pct, verbose = verbose, only.pos = only.pos,
max.cells.per.ident = max.cells.per.ident, random.seed = random.seed,
latent.vars = latent.vars, min.cells.feature = min.cells.feature,
min.cells.group = min.cells.group, pseudocount.use = pseudocount.use,
fc.results = fc.results, densify = densify, ...)

5: FindMarkers(object = data.use, slot = data.slot, counts = counts,
cells.1 = cells.1, cells.2 = cells.2, features = features,
logfc.threshold = logfc.threshold, test.use = test.use, min.pct = min.pct,
min.diff.pct = min.diff.pct, verbose = verbose, only.pos = only.pos,
max.cells.per.ident = max.cells.per.ident, random.seed = random.seed,
latent.vars = latent.vars, min.cells.feature = min.cells.feature,
min.cells.group = min.cells.group, pseudocount.use = pseudocount.use,
fc.results = fc.results, densify = densify, ...)

4: FindMarkers.SCTAssay(object = data.use, slot = slot, cells.1 = cells$cells.1,
cells.2 = cells$cells.2, features = features, logfc.threshold = logfc.threshold,
test.use = test.use, min.pct = min.pct, min.diff.pct = min.diff.pct,
verbose = verbose, only.pos = only.pos, max.cells.per.ident = max.cells.per.ident,
random.seed = random.seed, latent.vars = latent.vars, min.cells.feature = min.cells.feature,
min.cells.group = min.cells.group, mean.fxn = mean.fxn, base = base,
fc.name = fc.name, densify = densify, norm.method = norm.method,
...)

3: FindMarkers(object = data.use, slot = slot, cells.1 = cells$cells.1,
cells.2 = cells$cells.2, features = features, logfc.threshold = logfc.threshold,
test.use = test.use, min.pct = min.pct, min.diff.pct = min.diff.pct,
verbose = verbose, only.pos = only.pos, max.cells.per.ident = max.cells.per.ident,
random.seed = random.seed, latent.vars = latent.vars, min.cells.feature = min.cells.feature,
min.cells.group = min.cells.group, mean.fxn = mean.fxn, base = base,
fc.name = fc.name, densify = densify, norm.method = norm.method,
...)

2: FindMarkers.Seurat(immune.combined.sct, assay = "SCT", ident.1 = "B_STIM",
ident.2 = "B_CTRL", test.use = "MAST")

1: FindMarkers(immune.combined.sct, assay = "SCT", ident.1 = "B_STIM",
ident.2 = "B_CTRL", test.use = "MAST")

amcdavid · 2022-10-04T23:17:19Z

Yeah, this is a Seurat bug. It probably just needs to explicitly pass the norm.method argument to the FindMarkers, so it doesn't get slurped up by the ....

I see you opened an issue there, thanks: satijalab/seurat#6489

behrouzsh · 2022-10-05T03:23:29Z

Dear @amcdavid, even though there is another thread for this question, there is no answer for how to solve the issue, it is only said that there is a bug. Will you please be more specific? I am working on a spatial transcriptomic data and the available assays are "Spatial" and "SCT". MAST works for "Spatial" assay but generates the error for SCT. What should be the argument for norm.method to pass to FindMarkers for SCT assay? What is your suggested solution?
Thanks,

lucygarner · 2022-10-18T12:41:56Z

I am having the same issue. It would be great to get help with this.

MadsE10 · 2022-11-01T17:37:51Z

I had the same error and installing an older version of Seurat resolved it for me.

I previously had Seurat 4.2, but installed Seurat 4.1.1 then everything began to work normally again.

#just to check what version you are actually using
packageVersion('Seurat')

library(remotes)

remotes::install_version("Seurat", version = "4.1.1")

lopes36363 · 2023-03-06T00:32:47Z

If you delete SCT.model list from the seurat object, it suddenly works. I could remove it by:

sne[["LOL"]] <- CreateAssayObject(counts = GetAssayData(sne, assay="SCT", slot="data")) #Copy "SCT data" to "LOL counts" and "LOL data"
sne <- SetAssayData(sne, assay = "LOL", slot = "scale.data", new.data = GetAssayData(sne, assay="SCT", slot="scale.data")); gc() #"Copy "SCT scale.data" to "LOL scale.data"
DefaultAssay(sne) <- "LOL"
sne <- DietSeurat(sne, assays = "LOL", counts=T, data=T, scale.data=T)
sne <- RenameAssays(sne, LOL = 'SCT')

lwhitmore · 2023-03-13T23:48:55Z

Hello, I have downloaded the latest version of MAST 1.25 but am still getting the following error Seurat is version 4.3:
Error in .nextMethod(.Object = .Object, ... = ..., design = design) :
invalid name for slot of class “BayesGLMlike”: norm.method
Below is my command:
DEgenes <- FindMarkers(combined,
assay = "SCT",
slot="data",
ident.1 = ident_1,
ident.2 = ident_2,
test.use = "MAST",
logfc.threshold = 0,
min.pct = 0
# recorrect_umi = FALSE
)
Any help with this would be greatly appreciated

amcdavid · 2023-03-13T23:58:28Z

@lwhitmore Please provide a reproducible example that I can run to see if there's anything I can do about this issue MAST. The bug seems to be in Seurat, so you might open / comment on an issue there.

lwhitmore · 2023-03-14T05:37:02Z

@amcdavid My apologies I just needed to restart my R session apparently its working great! Again sorry to re-open this issue.

amcdavid mentioned this issue Oct 4, 2022

invalid name for slot of class "BayesGLMLike" #121

Closed

amcdavid mentioned this issue Nov 29, 2022

FindAllMarkers issue #176

Closed

amcdavid closed this as completed in 16e90b4 Nov 30, 2022

zhangsiyu1997 mentioned this issue Dec 4, 2022

"Warning:" for using "MAST" to find markers satijalab/seurat#6750

Closed

amcdavid reopened this Mar 13, 2023

amcdavid closed this as completed Mar 14, 2023

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

bug: MAST returns error after Seurat SCTransform, v2 regularization #175

bug: MAST returns error after Seurat SCTransform, v2 regularization #175

grhogg commented Sep 30, 2022

amcdavid commented Sep 30, 2022 via email

grhogg commented Sep 30, 2022

amcdavid commented Oct 4, 2022 •

edited

Loading

behrouzsh commented Oct 5, 2022

lucygarner commented Oct 18, 2022

MadsE10 commented Nov 1, 2022

lopes36363 commented Mar 6, 2023

lwhitmore commented Mar 13, 2023

amcdavid commented Mar 13, 2023

lwhitmore commented Mar 14, 2023

bug: MAST returns error after Seurat SCTransform, v2 regularization #175

bug: MAST returns error after Seurat SCTransform, v2 regularization #175

Comments

grhogg commented Sep 30, 2022

amcdavid commented Sep 30, 2022 via email

grhogg commented Sep 30, 2022

amcdavid commented Oct 4, 2022 • edited Loading

behrouzsh commented Oct 5, 2022

lucygarner commented Oct 18, 2022

MadsE10 commented Nov 1, 2022

lopes36363 commented Mar 6, 2023

lwhitmore commented Mar 13, 2023

amcdavid commented Mar 13, 2023

lwhitmore commented Mar 14, 2023

amcdavid commented Oct 4, 2022 •

edited

Loading