Merge branch 'main' into fix-average-batch-r2-celltype

Author: ghar1821

CHANGELOG.md

@@ -241,5 +241,14 @@
 * Update flowsom mapping similarity so we subset to just markers to correct, and lisi to remove control samples
   and unlabelled cells (PR #119).
 
+* Point `scripts/run_benchmark/wehi_hpc/run_full_hpc.sh` at `build/main` instead of
+  `build/update_ilisi`, and label the seqera full run as `full` instead of `test_subset` (PR #129).
+
+* Fix `average_batch_r2` and `flowsom_mapping_similarity` writing `metric_ids` and `metric_values`
+  as scalars instead of lists, as required by `file_score.yaml` (PR #130).
+
+* Clean up stale mock parameters and dead code (PR #133).
+
+
 * Fix bug in `average_batch_r2` where the R2 was computed on all cell types of a donor at once
   instead of on each cell type separately (PR #127).

_viash.yaml

@@ -38,7 +38,7 @@ description: |
 references: {}
 
 info:
-  image: The name of the image file to use for the component on the website.
+  image: thumbnail.svg
   test_resources:
     - type: s3
       path: s3://openproblems-data/resources_test/task_cyto_batch_integration/

scripts/run_benchmark/run_full_seqeracloud.sh

@@ -28,4 +28,4 @@ tw launch https://github.com/openproblems-bio/task_cyto_batch_integration.git \
   --params-file /tmp/params.yaml \
   --entry-name auto \
   --config common/nextflow_helpers/labels_tw.config \
-  --labels task_cyto_batch_integration,test_subset
+  --labels task_cyto_batch_integration,full

scripts/run_benchmark/wehi_hpc/run_full_hpc.sh

@@ -21,7 +21,7 @@ publish_dir: "$publish_dir"
 HERE
 
 tw launch https://github.com/openproblems-bio/task_cyto_batch_integration.git \
-  --revision build/update_ilisi \
+  --revision build/main \
   --pull-latest \
   --main-script target/nextflow/workflows/run_benchmark/main.nf \
   --workspace 80689470953249 \

src/control_methods/shuffle_integration_within_batch/script.py

@@ -4,10 +4,14 @@
 
 ## VIASH START
 par = {
-    "input_unintegrated": "resources_test/task_cyto_batch_integration/mouse_spleen_flow_cytometry_subset/unintegrated_censored.h5ad",
-    "output": "output.h5ad",
+    "input_unintegrated": "resources_test/task_cyto_batch_integration/mouse_spleen_flow_cytometry_subset/unintegrated.h5ad",
+    "output_integrated_split1": "resources_test/task_cyto_batch_integration/mouse_spleen_flow_cytometry_subset/control_integrated_split1.h5ad",
+    "output_integrated_split2": "resources_test/task_cyto_batch_integration/mouse_spleen_flow_cytometry_subset/control_integrated_split2.h5ad",
+}
+meta = {
+    "name": "shuffle_integration_within_batch",
+    "resources_dir": "src/control_methods",
 }
-meta = {"name": "shuffle_integration_within_batch"}
 ## VIASH END
 
 print("Importing helper functions", flush=True)

src/data_processors/process_dataset/script.py

@@ -7,7 +7,7 @@
     'input': 'resources_test/task_cyto_batch_integration/mouse_spleen_flow_cytometry_subset/common_dataset.h5ad',
     'output_censored_split1': 'resources_test/task_cyto_batch_integration/mouse_spleen_flow_cytometry_subset/censored_split1.h5ad',
     'output_censored_split2': 'resources_test/task_cyto_batch_integration/mouse_spleen_flow_cytometry_subset/censored_split2.h5ad',
-    'output_validation': 'resources_test/task_cyto_batch_integration/mouse_spleen_flow_cytometry_subset/validation.h5ad'
+    'output_unintegrated': 'resources_test/task_cyto_batch_integration/mouse_spleen_flow_cytometry_subset/unintegrated.h5ad'
 }
 meta = {
     'resources_dir': 'target/executable/data_processors/process_dataset',

src/methods/gaussnorm/script.R

@@ -4,12 +4,13 @@ library(flowStats)
 
 ## VIASH START
 par <- list(
-  input = "resources_test/task_cyto_batch_integration/cyto_spleen_subset/unintegrated_censored.h5ad",
+  input = "resources_test/task_cyto_batch_integration/mouse_spleen_flow_cytometry_subset/censored_split1.h5ad",
   output = "output.h5ad"
 )
 meta <- list(
-  name = "gaussNorm",
-  temp_dir: '/tmp'
+  name = "gaussnorm",
+  temp_dir = "/tmp",
+  resources_dir = "src/utils"
 )
 ## VIASH END
 

src/metrics/average_batch_r2/script.py

@@ -16,7 +16,7 @@
 ## VIASH END
 
 sys.path.append(meta["resources_dir"])
-from helper import batch_r2, concat_paired_samples, fit_r2
+from helper import batch_r2
 from helper_functions import (
     get_obs_var_for_integrated,
     remove_unlabelled,
@@ -40,10 +40,7 @@
 integrated_s2 = subset_nocontrols(integrated_s2)
 integrated_s2 = subset_markers_tocorrect(integrated_s2)
 
-print(
-    integrated_s1.obs, integrated_s2.obs, flush=True
-)  ### Debugging line, can be removed later
-print("Computing average_batch_r2 global", flush=True)
+print("Computing average_batch_r2 per cell type", flush=True)
 
 donor_list = integrated_s1.obs["donor"].unique()
 
@@ -84,8 +81,8 @@
     uns={
         "dataset_id": integrated_s1.uns["dataset_id"],
         "method_id": integrated_s1.uns["method_id"],
-        "metric_ids": "average_batch_r2_ct",
-        "metric_values": average_batch_r2_ct,
+        "metric_ids": ["average_batch_r2_ct"],
+        "metric_values": [average_batch_r2_ct],
         "r2_collection_ct": r2_collection_ct,
     }
 )

src/metrics/flowsom_mapping_similarity/script.R

@@ -144,8 +144,8 @@ output <- anndata::AnnData(
   uns = list(
     dataset_id = integrated_s1$uns$dataset_id,
     method_id = integrated_s1$uns$method_id,
-    metric_ids = "flowsom_mean_mapping_similarity",
-    metric_values = fs_mapping_similarity_avg,
+    metric_ids = list("flowsom_mean_mapping_similarity"),
+    metric_values = list(fs_mapping_similarity_avg),
     fsom_absdiff_by_donor_refsplit = fsom_absdiff_by_donor_refsplit,
     fsom_parameters = list(
       "xdim" = grid_xdim,

src/utils/helper_functions.R

@@ -150,43 +150,3 @@ remove_unlabelled <- function(adata) {
         c("unlabelled", "unlabeled")
     adata[!is_unlabelled, ]
 }
-
-#' Subsets the anndata object in a stratified manner
-#' with 'cell type' and 'sample' as strata.
-#'
-#' @param adata AnnData object
-#' @param frac numeric, fraction of cells to keep for each cell type
-#' @param seed numeric, seed for reproducibility
-#' @param anndatar logical, whether the input is anndataR object or not
-#' @return AnnData object with only the markers to correct
-subset_by_celltype <- function(adata, frac = 0.5, seed = 1, anndatar = TRUE) {
-  set.seed(seed)
-  
-  obs <- adata$obs
-  obs$cell_id <- rownames(obs)
-  obs$.row <- seq_len(nrow(obs))   # original order
-  
-  keep_ids <- obs %>%
-    group_by(cell_type, sample) %>%
-    slice_sample(prop = frac) %>%
-    ungroup() %>%
-    arrange(.row) %>%     # restore original order
-    pull(cell_id)
-  
-  if (anndatar == TRUE){
-    keep_idx <- match(keep_ids, adata$obs_names)
-    
-    adata_sub <- anndataR::AnnData(
-      X   = NULL,
-      obs = adata$obs[keep_idx, , drop = FALSE],
-      var = adata$var,
-      uns = adata$uns,
-      layers = list(
-        "integrated" = adata$layers$integrated[keep_idx, , drop = FALSE]
-      )
-    )
-  } else{
-    adata_sub <- adata[keep_ids, ]
-  }
-}
-