updated version of plotting function #199
Conversation
plus usability and scalability tabs + input metrics files
There was a problem hiding this comment.
I was able to run these and get some plots that look correct but the functions are quite complex so it's possible I have overlooked a bug somewhere.
Some general notes:
- It took we a while to work out how to run these, mostly that I had to set the working directory to where the scripts are. Maybe it would be good to have some documentation about how to use the functions? There is an empty README in this directory that could be a good place.
- Not sure if it would be good to have some control over where the output files are saved?
- These functions could maybe do with a bit of a tidy up which could make them easier to follow. Probably not a priority at this stage though.
- Do we need both the RNA metrics CSVs or just the most recent one? Might be better for git if they are named without the dates as well (but not entirely sure about that).
|
I have updated my plotting functions on this branch as well so @martaint you might need to do a |
| best_methods <- list( | ||
| RNA = c( | ||
| "Scanorama_embed_HVG_scaled", | ||
| "scANVI*_embed_HVG_unscaled", | ||
| "scGen*_gene_HVG_unscaled", | ||
| "fastMNN_embed_HVG_unscaled", | ||
| "BBKNN_graph_HVG_unscaled", | ||
| "Scanorama_gene_HVG_scaled", | ||
| "scVI_embed_HVG_unscaled", | ||
| "Seurat v3 RPCA_gene_HVG_unscaled", | ||
| "Harmony_embed_HVG_unscaled", | ||
| "Seurat v3 CCA_gene_HVG_unscaled", | ||
| "fastMNN_gene_HVG_unscaled", | ||
| "Conos_graph_FULL_unscaled", | ||
| "ComBat_gene_HVG_unscaled", | ||
| "MNN_gene_HVG_unscaled", | ||
| "trVAE_embed_HVG_unscaled", | ||
| "DESC_embed_FULL_unscaled", | ||
| "LIGER_embed_HVG_unscaled", | ||
| "SAUCIE_embed_HVG_scaled", | ||
| "SAUCIE_gene_HVG_scaled" | ||
| ), | ||
| ATAC = c( | ||
| "Harmony_embed", | ||
| "scANVI*_embed", | ||
| "LIGER_embed", | ||
| "scVI_embed", | ||
| "ComBat_gene", | ||
| "scGen*_gene", | ||
| "Seurat v3 RPCA_gene", | ||
| "Seurat v3 CCA_gene", | ||
| "Scanorama_embed", | ||
| "MNN_gene", | ||
| "trVAE_embed", | ||
| "fastMNN_embed", | ||
| "BBKNN_graph", | ||
| "Scanorama_gene", | ||
| "fastMNN_gene", | ||
| "DESC_embed", | ||
| "SAUCIE_embed", | ||
| "SAUCIE_gene", | ||
| "Conos_graph" | ||
| ) | ||
| ) |
There was a problem hiding this comment.
I should probably not hard code these but not sure if there is an easy way to get them. @martaint does/could your best methods functions output to best versions of each method? Would just need to be like what's here (method, output, features, scaling). Then I could make the best scatter functions more usable by somebody else.
| methods_pal <- c( | ||
| "BBKNN" = "#5A5156", | ||
| "Conos" = "#E4E1E3", | ||
| "trVAE" = "#F6222E", | ||
| "scVI" = "#FE00FA", | ||
| "ComBat" = "#16FF32", | ||
| "Harmony" = "#3283FE", | ||
| "LIGER" = "#FEAF16", | ||
| "Scanorama" = "#B00068", | ||
| "Seurat v3 CCA" = "#1CFFCE", | ||
| "Seurat v3 RPCA" = "#90AD1C", | ||
| "MNN" = "#2ED9FF", | ||
| "fastMNN" = "#DEA0FD", | ||
| "scGen*" = "#AA0DFE", | ||
| "scANVI*" = "#F8A19F", | ||
| "DESC" = "#325A9B", | ||
| "SAUCIE" = "#C4451C", | ||
| "Unintegrated" = "#66B0FF" | ||
| ) |
There was a problem hiding this comment.
This maybe shouldn't be hardcoded either...
There was a problem hiding this comment.
Could just be added to a separate mapping file that is loaded, like the tool name conversions above.
There was a problem hiding this comment.
This is one of mine. I'll fix it after this PR is merged.
| pancreas = "Pancreas", | ||
| lung_atlas = "Lung", | ||
| immune_cell_hum = "Immune (human)", | ||
| immune_cell_hum_mou = "Immune (human/mouse)", | ||
| mouse_brain = "Mouse brain", | ||
| simulations_1_1 = "Sim 1", | ||
| simulations_2 = "Sim 2", | ||
| mouse_brain_atac_genes_large = "ATAC large (genes)", | ||
| mouse_brain_atac_peaks_large = "ATAC large (peaks)", | ||
| mouse_brain_atac_windows_large = "ATAC large (windows)", | ||
| mouse_brain_atac_genes_small = "ATAC small (genes)", | ||
| mouse_brain_atac_peaks_small = "ATAC small (peaks)", | ||
| mouse_brain_atac_windows_small = "ATAC small (windows)" |
|
I also need to update my visualisation scripts but I think it will be easier to do it after this is merged (any maybe after they are moved to |
|
@martaint you can merge once you're happy with this... but please update the wording in the description a bit. |
| methods_pal <- c( | ||
| "BBKNN" = "#5A5156", | ||
| "Conos" = "#E4E1E3", | ||
| "trVAE" = "#F6222E", | ||
| "scVI" = "#FE00FA", | ||
| "ComBat" = "#16FF32", | ||
| "Harmony" = "#3283FE", | ||
| "LIGER" = "#FEAF16", | ||
| "Scanorama" = "#B00068", | ||
| "Seurat v3 CCA" = "#1CFFCE", | ||
| "Seurat v3 RPCA" = "#90AD1C", | ||
| "MNN" = "#2ED9FF", | ||
| "fastMNN" = "#DEA0FD", | ||
| "scGen*" = "#AA0DFE", | ||
| "scANVI*" = "#F8A19F", | ||
| "DESC" = "#325A9B", | ||
| "SAUCIE" = "#C4451C", | ||
| "Unintegrated" = "#66B0FF" | ||
| ) |
There was a problem hiding this comment.
Could just be added to a separate mapping file that is loaded, like the tool name conversions above.
|
Hello. I recently invested several hours trying to test / run the plotSingleAtlas.R function. It turned out to be a very slow process, and it required multiple manual installations, sometimes not described here (ggimage/dynutils/magick conflicts). I would like to test this PR to help debugging in case useful. Which is the branch I should clone locally in order to do it? A Thanks! |
|
the branch would be |
@martaint could you take a look at this? |
|
Where are we at with this? I'd like to tick off uploading a couple of my scripts but I think it's probably best if this is merged first. |
|
Took care of all comments! Thanks :) |
plus usability and scalability tabs + input metrics files