liana_tensor_c2c aborting R session

[yannk-lm]

Hello,

Thank you for developing such a useful package.

I'm trying to follow the liana_c2ctensor vignette. However, the liana_tensor_c2c function aborts my R session. I found out that it was the reticulate::import function causing this, even when I was setting the conda_env argument to the directory where I installed cell2cell (following their installation process).

This seems to be an issue on the reticulate end, but I really don't know how to proceed from there. Do you have any idea how to fix this?

Please find my R session below.

Thank you for your help, Yannick

R version 4.2.2 (2022-10-31 ucrt) Platform: x86_64-w64-mingw32/x64 (64-bit) Running under: Windows 10 x64 (build 19044)

Matrix products: default

locale: [1] LC_COLLATE=French_France.utf8 LC_CTYPE=French_France.utf8 LC_MONETARY=French_France.utf8 [4] LC_NUMERIC=C LC_TIME=French_France.utf8

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

other attached packages: [1] liana_0.1.11 basilisk_1.10.2 reticulate_1.28

loaded via a namespace (and not attached): [1] colorspace_2.1-0 rjson_0.2.21 ellipsis_0.3.2 scuttle_1.8.4
[5] circlize_0.4.15 bluster_1.8.0 XVector_0.38.0 BiocNeighbors_1.16.0
[9] GenomicRanges_1.50.2 GlobalOptions_0.1.2 clue_0.3-64 rstudioapi_0.14
[13] listenv_0.9.0 fansi_1.0.4 xml2_1.3.3 sparseMatrixStats_1.10.0
[17] codetools_0.2-19 doParallel_1.0.17 knitr_1.42 jsonlite_1.8.4
[21] cluster_2.1.4 png_0.1-8 readr_2.1.4 compiler_4.2.2
[25] httr_1.4.4 dqrng_0.3.0 backports_1.4.1 SeuratObject_4.1.3
[29] Matrix_1.5-3 fastmap_1.1.0 limma_3.54.1 cli_3.6.0
[33] later_1.3.0 BiocSingular_1.14.0 htmltools_0.5.4 prettyunits_1.1.1
[37] tools_4.2.2 rsvd_1.0.5 igraph_1.4.0 gtable_0.3.1
[41] glue_1.6.2 GenomeInfoDbData_1.2.9 dplyr_1.1.0 rappdirs_0.3.3
[45] Rcpp_1.0.10 Biobase_2.58.0 cellranger_1.1.0 vctrs_0.5.2
[49] progressr_0.13.0 DelayedMatrixStats_1.20.0 iterators_1.0.14 OmnipathR_3.6.0
[53] xfun_0.37 stringr_1.5.0 globals_0.16.2 beachmat_2.14.0
[57] rvest_1.0.3 irlba_2.3.5.1 lifecycle_1.0.3 statmod_1.5.0
[61] future_1.31.0 edgeR_3.40.2 zlibbioc_1.44.0 scales_1.2.1
[65] basilisk.utils_1.10.0 hms_1.1.2 MatrixGenerics_1.10.0 parallel_4.2.2
[69] SummarizedExperiment_1.28.0 RColorBrewer_1.1-3 SingleCellExperiment_1.20.0 ComplexHeatmap_2.14.0
[73] yaml_2.3.7 curl_5.0.0 ggplot2_3.4.1 stringi_1.7.12
[77] S4Vectors_0.36.1 foreach_1.5.2 ScaledMatrix_1.6.0 checkmate_2.1.0
[81] scran_1.26.2 BiocGenerics_0.44.0 filelock_1.0.2 BiocParallel_1.32.5
[85] shape_1.4.6 GenomeInfoDb_1.34.9 rlang_1.0.6 pkgconfig_2.0.3
[89] matrixStats_0.63.0 bitops_1.0-7 evaluate_0.20 lattice_0.20-45
[93] purrr_1.0.1 tidyselect_1.2.0 parallelly_1.34.0 logger_0.2.2
[97] magrittr_2.0.3 R6_2.5.1 IRanges_2.32.0 generics_0.1.3
[101] metapod_1.6.0 DelayedArray_0.24.0 pillar_1.8.1 withr_2.5.0
[105] RCurl_1.98-1.10 sp_1.6-0 tibble_3.1.8 future.apply_1.10.0
[109] dir.expiry_1.6.0 crayon_1.5.2 utf8_1.2.3 tzdb_0.3.0
[113] rmarkdown_2.20 GetoptLong_1.0.5 progress_1.2.2 locfit_1.5-9.7
[117] grid_4.2.2 readxl_1.4.2 digest_0.6.31 tidyr_1.3.0
[121] stats4_4.2.2 munsell_0.5.0

[dbdimitrov]

Hi @yannk-lm,

I tried to run it in an environment with R4.2 & reticulate_1.28 and it worked just fine on my PC, but it's a linux and I cannot test on Windows - it might be related to that. You could try to see if reticulate alone crashes your session as that's all that needs to load for the rest of the pipeline to work.

Hope this helps.

[dbdimitrov]

Another alternative would be to run liana and Tensor purely in Python: https://liana-py.readthedocs.io/en/latest/notebooks/liana_c2c.html

[yannk-lm]

Hello,

Since you mentioned that it could be a Windows issue, I ran R on WSL and reticulate::import is now functioning properly.

Thank you for your help, Yannick