Improvements based on reviewers comments

[dimalvovs]

Here is the reviewer's comment:

The provided source code does not include any documentation regarding how should the code by used, in what order, under which software requirements and includes hardcoded paths to the authors' machine, making it unusable and hard to scrutinize. Improving the code would enable reproduction of the results and broad use of the analytical approach since the paper is mainly about bringing forth a new apporoach.
Furthermore, I would recommend also depositing the source code in a openly accessibly, independent repository for archival such as Zenodo, which can be done easily by connecting the Github repositorie(s) to Zenodo.

Code documentation improvements

• Vignettes?

visium analysis scripts

• possibly non-existant variable Grade_pal

• are these comments needed?

  PanIN_carcinogeneisis_spatial_analysis/scripts/visium_analysis/PDAC_atlas_projection/00_PanIN_Custom_Functions.R

  Line 1031 in 52f6c9c

  # alpha is transparency of spots (I think, tbh don't remember right now), just leave that as default

  
  

  PanIN_carcinogeneisis_spatial_analysis/scripts/visium_analysis/PDAC_atlas_projection/00_PanIN_Custom_Functions.R

  Line 483 in 52f6c9c

  # if (do.identify) {

  PanIN_carcinogeneisis_spatial_analysis/scripts/visium_analysis/PDAC_atlas_projection/00_PanIN_Custom_Functions.R

  Line 724 in 52f6c9c

  # spot.size <- slot(object = image, name = "spot.radius")

  
  

  PanIN_carcinogeneisis_spatial_analysis/scripts/visium_analysis/PDAC_atlas_projection/00_PanIN_Custom_Functions.R

  Line 755 in 52f6c9c

  # ggplot2:::ggname("geom_spatial", gt)

• fix broken source paths, for example

  PanIN_carcinogeneisis_spatial_analysis/scripts/visium_analysis/PDAC_atlas_projection/01_Pre_processing_paired_cohort.R

  Line 26 in 52f6c9c

  source("..Custom_Functions.R")

  
  

  PanIN_carcinogeneisis_spatial_analysis/scripts/visium_analysis/PDAC_atlas_projection/02_Paired_cohort_visium_analysis.R

  Line 26 in 52f6c9c

  source("..Custom_Functions.R")

• folder structure and path updates
  

  PanIN_carcinogeneisis_spatial_analysis/scripts/visium_analysis/PDAC_atlas_projection/01_Pre_processing_paired_cohort.R

  Line 144 in 52f6c9c

  data_merged@reductions$harmony <- readRDS("..objects_and_metadata/harmony_reduction.rds")

  
  

  PanIN_carcinogeneisis_spatial_analysis/scripts/visium_analysis/PDAC_atlas_projection/01_Pre_processing_paired_cohort.R

  Line 163 in 52f6c9c

  data_merged$SCT_snn_res.0.5 <- readRDS("..objects_and_metadata/harmony_louvain.rds")

xenium analysis scripts

• do we need to store these comments?

  PanIN_carcinogeneisis_spatial_analysis/scripts/xenium_analysis/01_Load_Xenium_Data.Rmd

  Line 13 in 52f6c9c

  # new lib path for the most current version of Seurat that can use Xenium data

• should we export the conda env to allow reproducing with the same package versions?

  PanIN_carcinogeneisis_spatial_analysis/scripts/xenium_analysis/01_Load_Xenium_Data.Rmd

  Line 23 in 52f6c9c

  use_condaenv(condaenv = "panin_visium")

• if we use conda_env, do we need to specify the lib loc?

  PanIN_carcinogeneisis_spatial_analysis/scripts/xenium_analysis/01_Load_Xenium_Data.Rmd

  Line 18 in 52f6c9c

  library(sp, lib.loc = "~/FertigLab/Seurat_Xenium")

closing

• archive the version in Zenodo

[jmitchell81]

Xenium comments addressed in d17a538. The conda environment usage is removed because it was a holdover from an initial attempt to use Leiden clustering. The presented scripts use Louvain clustering.

[jmitchell81]

Commented out code removed in 091299b