carlosproca / autospill Goto Github PK

Currently autospill has a list of forbidden characters that might go beyond its actual needs, for instance I have tested removing some of them like /, and the process still works well AFAIK.

The issue is that the silent substitution of these characters means that the matrices produced have marker names that are different to what is originally present in the FCS / control file provided. While initially invisible to the user, once a compensation is tried with the matrix against FCS files that contain the same marker names as the original ones it will produce errors like:

Error: input_dir/my-file.fcs
The following parameters in the spillover matrix
 are not present in the flowFrame:
...
Execution halted

This happens at https://github.com/carlosproca/autospill/blob/master/R/read_marker.r#L67

I have circumvented this partly on the Galaxy module by giving the user the ability to remove desired characters from the forbidden list (at the risk of breaking the execution), but maybe autospill might want to explicitly warn the user that certain markers/dyes names have been changed in the process by listing the original and final names (with this the user could react my amending the matrices for instance to keep them coherent with the FCS files).

Hi there!

Thanks for this great tool. I'm currently packaging it for bioconda to use it within Galaxy, but for bioconda this needs to have:

• A formalised license (GPL, LGPL, MIT, or other; essentially a known license model that allows re-distribution)
• Format the version to the proper semantic versioning standard, in this case for the current version should be 0.1.0 if I'm not mistaken.

Happy to PR, but specially the first one is a decision that you need to make.

Thanks!

I tried running the example code in calculate_compensation_minimal.r for the AutoSpill - MM1 dataset. I can import the data using read.flow.control (at least without significant errors), but unfortunately the gate.flow.data results in the following error:

Error in mclapply(flow.control$sample, function(samp) do.gate(flow.control$expr.data.untr[flow.control$event.sample == : 'mc.cores' > 1 is not supported on Windows

A bit of googling tells me this is a known issue for mclapply in Windows. I think this means that currently this script cannot run on Windows.

Here is my session info:

R version 4.0.2 (2020-06-22)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 19041)

Matrix products: default

locale:
[1] LC_COLLATE=English_United States.1252 LC_CTYPE=English_United States.1252
[3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C
[5] LC_TIME=English_United States.1252

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

other attached packages:
[1] autospill_0.01.00

loaded via a namespace (and not attached):
[1] Rcpp_1.0.5 compiler_4.0.2 pillar_1.4.6 RColorBrewer_1.1-2
[5] cytolib_2.0.3 moments_0.14 tools_4.0.2 zlibbioc_1.34.0
[9] digest_0.6.25 dotCall64_1.0-0 gtable_0.3.0 lifecycle_0.2.0
[13] tibble_3.0.3 lattice_0.20-41 png_0.1-7 pkgconfig_2.0.3
[17] rlang_0.4.7 graph_1.66.0 rstudioapi_0.11 Rgraphviz_2.32.0
[21] parallel_4.0.2 spam_2.5-1 stringr_1.4.0 dplyr_1.0.2
[25] generics_0.0.2 vctrs_0.3.4 fields_11.4 maps_3.3.0
[29] stats4_4.0.2 grid_4.0.2 tidyselect_1.1.0 glue_1.4.2
[33] data.table_1.13.0 Biobase_2.48.0 R6_2.4.1 jpeg_0.1-8.1
[37] XML_3.99-0.5 sp_1.4-2 latticeExtra_0.6-29 ggplot2_3.3.2
[41] RProtoBufLib_2.0.0 purrr_0.3.4 deldir_0.1-28 magrittr_1.5
[45] scales_1.1.1 matrixStats_0.56.0 ellipsis_0.3.1 MASS_7.3-51.6
[49] BiocGenerics_0.34.0 colorspace_1.4-1 flowWorkspace_4.0.6 flowCore_2.0.1
[53] ncdfFlow_2.34.0 stringi_1.5.3 tripack_1.3-9.1 munsell_0.5.0
[57] RcppParallel_5.0.2 crayon_1.3.4

parallel::detectCores() [1] 8

Hi there!

I tried running the first dataset example with the calculate_compensation_paper.r script on a 2.5 GHz Dual core Intel Core i7. I see 4 R threads running for ~8 hours now (and is not done yet). Is that expected running time or I should be looking into something else in detail as this is too long for the MM1 dataset?

Thanks!

The autospill website version cannot be accessed. https://autospill.vib.be/
Is there a new address or the website version is permanently down?