Differences running Seurat's tutorial

[JMRC-08]

Hi Seurat Team

I'm new to analyzing scRNA-Seq data with Seurat package v.4.0.6. When I was trying to replicate the tutorial with the pbmc3k data I got differences with some plot results. Respect to JackStrawPlot (pbmc, dims = 1:15), some of the p-values of each PC are the same as in the figure shown in the tutorial, except for PC 3, 4, 6, 7, 11, 13, and 15. Here is the JackStraw plot I generated:

When I ran non-linear dimensional reduction (UMAP/tSNE) and plotted the results, I generated the following figure:

Several of the clusters are shifted from their locations in the figure from the tutorial:

What are the reasons for these observed differences in the location of the clusters?

I used the pbmc3k_tutorial.rmd pipeline , so there should not be any differences in the code. I am not very familiar with Python and R, so I suspect whatever I'm doing wrong has something to do with that. I am sorry if this is an obvious question, but I could not find a definitive answer, and would not want to make a silly mistake.

The output of sessionInfo():
R version 4.1.1 (2021-08-10)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS Big Sur 11.6.4

Matrix products: default
LAPACK: /Library/Frameworks/R.framework/Versions/4.1/Resources/lib/libRlapack.dylib

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

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

other attached packages:
[1] ggplot2_3.3.5 patchwork_1.1.1 SeuratObject_4.0.4 Seurat_4.0.6
[5] dplyr_1.0.7

loaded via a namespace (and not attached):
[1] nlme_3.1-153 matrixStats_0.61.0 spatstat.sparse_2.1-0
[4] RcppAnnoy_0.0.19 RColorBrewer_1.1-2 httr_1.4.2
[7] sctransform_0.3.3 tools_4.1.1 utf8_1.2.2
[10] R6_2.5.1 irlba_2.3.5 rpart_4.1-15
[13] KernSmooth_2.23-20 uwot_0.1.11 mgcv_1.8-37
[16] DBI_1.1.1 lazyeval_0.2.2 colorspace_2.0-2
[19] withr_2.5.0 tidyselect_1.1.1 gridExtra_2.3
[22] compiler_4.1.1 cli_3.2.0 plotly_4.10.0
[25] labeling_0.4.2 scales_1.1.1 lmtest_0.9-39
[28] spatstat.data_2.1-2 ggridges_0.5.3 pbapply_1.5-0
[31] goftest_1.2-3 stringr_1.4.0 digest_0.6.28
[34] spatstat.utils_2.3-0 pkgconfig_2.0.3 htmltools_0.5.2
[37] parallelly_1.30.0 fastmap_1.1.0 htmlwidgets_1.5.4
[40] rlang_1.0.2 rstudioapi_0.13 shiny_1.7.1
[43] farver_2.1.0 generics_0.1.0 zoo_1.8-9
[46] jsonlite_1.7.2 ica_1.0-2 magrittr_2.0.1
[49] Matrix_1.3-4 Rcpp_1.0.7 munsell_0.5.0
[52] fansi_0.5.0 abind_1.4-5 reticulate_1.22
[55] lifecycle_1.0.1 stringi_1.7.5 MASS_7.3-54
[58] Rtsne_0.15 plyr_1.8.6 grid_4.1.1
[61] parallel_4.1.1 listenv_0.8.0 promises_1.2.0.1
[64] ggrepel_0.9.1 crayon_1.4.1 deldir_1.0-6
[67] miniUI_0.1.1.1 lattice_0.20-45 cowplot_1.1.1
[70] splines_4.1.1 tensor_1.5 knitr_1.36
[73] pillar_1.6.3 igraph_1.2.11 spatstat.geom_2.3-1
[76] future.apply_1.8.1 reshape2_1.4.4 codetools_0.2-18
[79] leiden_0.3.9 glue_1.6.2 data.table_1.14.2
[82] png_0.1-7 vctrs_0.3.8 httpuv_1.6.5
[85] gtable_0.3.0 RANN_2.6.1 purrr_0.3.4
[88] spatstat.core_2.3-2 polyclip_1.10-0 tidyr_1.1.4
[91] scattermore_0.7 future_1.23.0 assertthat_0.2.1
[94] xfun_0.26 mime_0.12 xtable_1.8-4
[97] RSpectra_0.16-0 later_1.3.0 survival_3.2-13
[100] viridisLite_0.4.0 tibble_3.1.5 cluster_2.1.2
[103] globals_0.14.0 fitdistrplus_1.1-6 ellipsis_0.3.2
[106] ROCR_1.0-11

Your suggestions would be greatly appreciated.

[AuritaP]

I have the same results/problem. First I did exactly as the instructions says and in my system, even though it ran, I got a comment saying that "The default method for RunUMAP has changed from calling Python UMAP via reticulate to the R-native UWOT using the cosine metric. To use Python UMAP via reticulate, set umap.method to 'umap-learn' and metric to 'correlation'. So I thought maybe that is the reason for the discrepancy, so I re-run the whole thing using those parameters to load UMAP and still has the same discrepancy with slight different results. I would like to know if this is ok or not

[Ck-boo]

Hello everyone!

Complete rookie in bioinformatics, scRNA-seq analysis, R and all! I have a similar issue, but my plot looks like this:

Could someone explain why this is happening and if it can be fixed? Make it as simple as possible please, this is an entirely new world for me :D

Thank you,
Ck