Skip to content
This repository was archived by the owner on Oct 14, 2025. It is now read-only.

Commit e333130

Browse files
stemangiolastemangiola
authored
Merge pull request #112 from stemangiola/improve-figurse
Improve figures
2 parents 928e459 + 1db8409 commit e333130

File tree

3 files changed

+86
-86
lines changed

3 files changed

+86
-86
lines changed

‎R/dev.R‎

Lines changed: 67 additions & 67 deletions
Original file line numberDiff line numberDiff line change
@@ -145,7 +145,7 @@ update_unharmonised <- function(unharmonised_parquet_dir, ...){
145145
#' "/vast/projects/cellxgene_curated/splitted_DB2_anndata_scaled_0.2.1"
146146
#' )
147147
#' }
148-
hdf5_to_anndata = function(input_directory, output_directory){
148+
hdf5_to_anndata <- function(input_directory, output_directory){
149149
dir.create(output_directory, showWarnings = FALSE)
150150
# This is a quick utility script to convert the SCE files into AnnData format for use in Pythonlist.files("/vast/projects/RCP/human_cell_atlas/splitted_DB2_data", full.names = FALSE) |> purrr::walk(function(dir){
151151
basilisk::basiliskRun(fun = function(sce) {
@@ -185,72 +185,72 @@ hdf5_to_anndata = function(input_directory, output_directory){
185185
}, env = zellkonverter::zellkonverterAnnDataEnv())
186186
}
187187

188-
#' Converts a series of H5-serialized Seurat to AnnData
189-
#' @param input_directory A character scalar. The path to a directory containing one or more
190-
#' directories created by [SeuratDisk::SaveH5Seurat()].
191-
#' @param output_directory A character scalar. The path to a directory in which to save the
192-
#' created anndata files.
193-
#' @keywords internal
194-
#' @return A character vector of the newly-created anndata files
195-
#' @examples
196-
#' \donttest{
197-
#' h5seurat_to_anndata(
198-
#' "/vast/projects/cellxgene_curated/splitted_DB2_data_0.2.1",
199-
#' "/vast/projects/cellxgene_curated/splitted_DB2_anndata_0.2.1"
200-
#' )
201-
#' h5seurat_to_anndata(
202-
#' "/vast/projects/cellxgene_curated/splitted_DB2_data_scaled_0.2.1",
203-
#' "/vast/projects/cellxgene_curated/splitted_DB2_anndata_scaled_0.2.1"
204-
#' )
205-
#' }
206-
h5seurat_to_anndata = function(input_directory, output_directory, assays = "RNA"){
207-
208-
# Check if package is loaded
209-
if(!"SeuratDisk" %in% (.packages()))
210-
stop("CuratedCellAtlas says: please manually load the SeuratDisk package first. Execute `library(SeuratDisk)`")
211-
212-
213-
dir.create(output_directory, showWarnings = FALSE)
214-
# This is a quick utility script to convert the SCE files into AnnData format for use in Pythonlist.files("/vast/projects/RCP/human_cell_atlas/splitted_DB2_data", full.names = FALSE) |> purrr::walk(function(dir){
215-
basilisk::basiliskRun(fun = function(sce) {
216-
dir(input_directory, full.names = TRUE) |>
217-
purrr::map_chr(function(seurat_file){
218-
cli::cli_alert_info("Processing {seurat_file}.")
219-
prefix <- basename(seurat_file)
220-
out_path <- glue::glue("{prefix}.h5ad") |>
221-
file.path(output_directory, name=_)
222-
223-
if (file.exists(out_path)) {
224-
cli::cli_alert_info("{out_path} already exists. Skipping")
225-
}
226-
else {
227-
sce <-
228-
LoadH5Seurat(seurat_file, assays = assays) |>
229-
Seurat::as.SingleCellExperiment()
230-
231-
single_column <- length(colnames(sce)) == 1
232-
if (single_column){
233-
# Hack, so that single-column SCEs will convert
234-
# correctly
235-
cli::cli_alert_info(
236-
"{seurat_file} has only 1 column. Duplicating column."
237-
)
238-
sce <- cbind(sce, sce)
239-
single_column <- TRUE
240-
}
241-
ad <- zellkonverter::SCE2AnnData(sce)
242-
if (single_column){
243-
# Remove the duplicate column
244-
sce$X <- sce$X[1]
245-
}
246-
# TODO: customize chunking here, when anndata supports it
247-
# (see https://github.com/scverse/anndata/issues/961)
248-
ad$write_h5ad(out_path)
249-
}
250-
out_path
251-
}, .progress = "Converting files")
252-
}, env = zellkonverter::zellkonverterAnnDataEnv())
253-
}
188+
# Converts a series of H5-serialized Seurat to AnnData
189+
# @param input_directory A character scalar. The path to a directory containing one or more
190+
# directories created by [SeuratDisk::SaveH5Seurat()].
191+
# @param output_directory A character scalar. The path to a directory in which to save the
192+
# created anndata files.
193+
# @keywords internal
194+
# @return A character vector of the newly-created anndata files
195+
# @noRd
196+
# @examples
197+
# \donttest{
198+
# h5seurat_to_anndata(
199+
# "/vast/projects/cellxgene_curated/splitted_DB2_data_0.2.1",
200+
# "/vast/projects/cellxgene_curated/splitted_DB2_anndata_0.2.1"
201+
# )
202+
# h5seurat_to_anndata(
203+
# "/vast/projects/cellxgene_curated/splitted_DB2_data_scaled_0.2.1",
204+
# "/vast/projects/cellxgene_curated/splitted_DB2_anndata_scaled_0.2.1"
205+
# )
206+
# }
207+
# h5seurat_to_anndata <- function(input_directory, output_directory, assays = "RNA"){
208+
#
209+
# # Check if package is loaded
210+
# if(!"SeuratDisk" %in% (.packages()))
211+
# stop("CuratedCellAtlas says: please manually load the SeuratDisk package first. Execute `library(SeuratDisk)`")
212+
#
213+
#
214+
# dir.create(output_directory, showWarnings = FALSE)
215+
# # This is a quick utility script to convert the SCE files into AnnData format for use in Pythonlist.files("/vast/projects/RCP/human_cell_atlas/splitted_DB2_data", full.names = FALSE) |> purrr::walk(function(dir){
216+
# basilisk::basiliskRun(fun = function(sce) {
217+
# dir(input_directory, full.names = TRUE) |>
218+
# purrr::map_chr(function(seurat_file){
219+
# cli::cli_alert_info("Processing {seurat_file}.")
220+
# prefix <- basename(seurat_file)
221+
# out_path <- glue::glue("{prefix}.h5ad") |>
222+
# file.path(output_directory, name=_)
223+
#
224+
# if (file.exists(out_path)) {
225+
# cli::cli_alert_info("{out_path} already exists. Skipping")
226+
# }
227+
# else {
228+
# sce <- SeuratDisk::LoadH5Seurat(seurat_file, assays = assays) |>
229+
# Seurat::as.SingleCellExperiment()
230+
#
231+
# single_column <- length(colnames(sce)) == 1
232+
# if (single_column){
233+
# # Hack, so that single-column SCEs will convert
234+
# # correctly
235+
# cli::cli_alert_info(
236+
# "{seurat_file} has only 1 column. Duplicating column."
237+
# )
238+
# sce <- cbind(sce, sce)
239+
# single_column <- TRUE
240+
# }
241+
# ad <- zellkonverter::SCE2AnnData(sce)
242+
# if (single_column){
243+
# # Remove the duplicate column
244+
# sce$X <- sce$X[1]
245+
# }
246+
# # TODO: customize chunking here, when anndata supports it
247+
# # (see https://github.com/scverse/anndata/issues/961)
248+
# ad$write_h5ad(out_path)
249+
# }
250+
# out_path
251+
# }, .progress = "Converting files")
252+
# }, env = zellkonverter::zellkonverterAnnDataEnv())
253+
# }
254254

255255
#' Makes a "downsampled" metadata file that only contains the minimal data
256256
#' needed to run the vignette.

man/dir_to_anndata.Rd renamed to man/hdf5_to_anndata.Rd

Lines changed: 7 additions & 7 deletions
Some generated files are not rendered by default. Learn more about customizing how changed files appear on GitHub.

‎vignettes/Introduction.Rmd‎

Lines changed: 12 additions & 12 deletions
Original file line numberDiff line numberDiff line change
@@ -215,7 +215,7 @@ single_cell_counts |> HDF5Array::saveHDF5SummarizedExperiment("single_cell_count
215215

216216
We can gather all CD14 monocytes cells and plot the distribution of HLA-A across all tissues
217217

218-
```{r}
218+
```{r, results='hide'}
219219
suppressPackageStartupMessages({
220220
library(ggplot2)
221221
})
@@ -251,7 +251,13 @@ counts |>
251251
theme(axis.text.x = element_text(angle = 60, vjust = 1, hjust = 1)) +
252252
xlab("Disease") +
253253
ggtitle("HLA-A in CD14 monocytes by disease")
254+
```
255+
256+
```{r echo=FALSE}
257+
find_figure("HLA_A_disease_plot.png") |> knitr::include_graphics()
258+
```
254259

260+
```{r, results='hide'}
255261
# Plot by tissue
256262
counts |>
257263
dplyr::with_groups(tissue_harmonised, ~ .x |> dplyr::mutate(median_count = median(`HLA.A`, rm.na=TRUE))) |>
@@ -269,16 +275,8 @@ counts |>
269275
ggtitle("HLA-A in CD14 monocytes by tissue") +
270276
theme(legend.position = "none")
271277
```
272-
273-
```{r}
274-
counts |>
275-
ggplot(aes( disease, `HLA.A`,color = file_id)) +
276-
geom_jitter(shape=".") +
277-
theme_bw() +
278-
theme(
279-
axis.text.x = element_text(angle = 60, vjust = 1, hjust = 1),
280-
legend.position = "none"
281-
)
278+
```{r echo=FALSE}
279+
find_figure("HLA_A_tissue_plot.png") |> knitr::include_graphics()
282280
```
283281

284282
```{r}
@@ -299,7 +297,9 @@ metadata |>
299297
axis.text.x = element_text(angle = 60, vjust = 1, hjust = 1),
300298
legend.position = "none"
301299
) +
302-
geom_jitter(shape=".")
300+
geom_jitter(shape=".") +
301+
xlab("Tissue") +
302+
ggtitle("HLA-A in nk cells by tissue")
303303
```
304304

305305
## Obtain Unharmonised Metadata

0 commit comments

Comments
 (0)