docs: fix typo in transcript_assign

docs/computational/generate_expression_matrix.md

@@ -111,7 +111,7 @@ Then, you can create a single file containing the transcriptomic information agg
           --sample-id openst_demo_sample \
           transcript_assign \
           --spatial-key obsm/spatial_pairwise_aligned_fine \
-          --image-key uns/spatial_pairwise_aligned/staining_image_transformed
+          --mask-in uns/spatial_pairwise_aligned/staining_image_transformed
      ```
 
 === "From semiautomatic alignment"
@@ -122,7 +122,7 @@ Then, you can create a single file containing the transcriptomic information agg
           --sample-id openst_demo_sample \
           transcript_assign \
           --spatial-key obsm/spatial_manual_fine \
-          --image-key uns/spatial_pairwise_aligned/staining_image_transformed
+          --mask-in uns/spatial_pairwise_aligned/staining_image_transformed
      ```
 
 === "From manual alignment"

examples/example_workflow.sh

@@ -1,82 +1,29 @@
-#!/bin/bash
+# 0. cd into a root spacemake folder (i.e., containing project_df.csv and config.yaml)
+# cd ...
 
-# author: Daniel Leon-Perinan (Rajewsky Lab, 2023)
-# this file uses openst to preprocess STS and imaging data, 
-# yields a single-cell h5ad file (UMI aggregated by segmentation mask)
+# 1. stitching of the spatial dge
+openst from_spacemake --project-id fc_sts_76 --sample-id fc_sts_76_3 spatial_stitch --tile-coordinates fc_2_coordinate_system.csv
 
-# placeholders:
-# <path> = path where results from this script will be saved
-# <path_image> = path to the folder containing the corresponding staining image
-# <path_spacemake_dge> = spacemake-generated path to the dge files of this sample
-# <id> = the sample id, as in spacemake
+# 2. copy images into raw directory
 
-### STEP 0 ###
-# activate the openst environment
-# e.g., conda activate openst
+# 3. stitching of the images
+# TODO: write tutorial for how to download ImageJ
+openst from_spacemake --project-id fc_sts_76 --sample-id fc_sts_76_3 image_stitch --imagej-bin ~/Fiji.app/ImageJ-linux64 --microscope keyence
 
-### STEP 1 ###
-# stitching of the individual tiles with a coordinate system
-openst spatial_stitch \
-    --tiles <path_spacemake_dge>/dge.all.polyA_adapter_trimmed.mm_included.spatial_beads_*.h5ad \
-    --tile-coordinates <path>/novaseq_S4_coordinate_system_FC_010.csv \
-    --output <path>/<id>_stitched_spots.h5ad \
-    --join-output 'outer' \
-    --tile-id-regex '(L[1-4]_tile_[1-2][0-7][0-9][0-9])'
+# 4. merge modalities
+openst from_spacemake --project-id fc_sts_76 --sample-id fc_sts_76_3 merge_modalities
 
-### STEP 2 ###
-# pairwise alignment of the coordinates
-# we only run coarse alignment
-# this will integrate the staining image into the file (nice for segmentation)
-openst pairwise_aligner \
-    --image-in <path_image>/Image_Stitched_Composite.tif \
-    --h5-in <path>/<id>_stitched_spots.h5ad \
-    --h5-out <path>/<id>_stitched_spots_aligned.h5ad \
-    --save-image-in-h5 \
-    --feature-matcher 'LoFTR' \
-    --ransac-coarse-max-trials 1000 \
-    --ransac-fine-max-trials 1000 \
-    --n-threads 8 \
-    --ransac-coarse-residual-threshold 2 \
-    --ransac-coarse-min-samples 2 \
-    --only-coarse \
-    --device cuda \
-    --metadata-out <path>/<id>_alignment_metadata.json \
-    --threshold-counts-coarse 2
+# 5.A pairwise alignment, automatic
+openst from_spacemake --project-id fc_sts_76 --sample-id fc_sts_76_3 pairwise_aligner --device cuda --mask-tissue --only-coarse
 
-### STEP 3 ###
-# requires manual input from the user!
-# we run the refinement with the gui
-# if a display is not set from ssh session, run ssh -X <host>, then e.g.,
-# export DISPLAY=localhost:10.0
-openst manual_pairwise_aligner_gui
+# 5.B pairwise alignment, manual
+# we load the GUI, then select points for coarse and fine registration. Do not forget to save the keypoints.json file!
+openst from_spacemake --project-id fc_sts_76 --sample-id fc_sts_76_3 manual_pairwise_aligner --spatial-key obsm/spatial --image-key uns/spatial/staining_image
+# we apply the transform to the data, per tile (cannot be done in GUI)
+openst from_spacemake --project-id fc_sts_76 --sample-id fc_sts_76_3 apply_transform --keypoints keypoints.json --spatial-key-in obsm/spatial_pairwise_aligned_coarse --spatial-key-out obsm/spatial_pairwise_aligned_fine --per-tile
 
-### STEP 4 ###
-# transfer the refined keypoints to the fine aligned data
-openst manual_pairwise_aligner \
-    --keypoints-json <path>/<id>_keypoints.json \
-    --h5-in <path>/<id>_stitched_spots_aligned.h5ad \
-    --fine
+# 6. segmentation
+openst from_spacemake --project-id fc_sts_76 --sample-id fc_sts_76_3 segment --device cuda --model HE_cellpose_rajewsky
 
-### STEP 5 ###
-# calculate a segmentation mask for the aligned staining image
-# optionally, we apply a dilation of 10px (increases UMIs, but might increase background signal)
-openst segment \
-    --adata <path>/<id>_stitched_spots_aligned.h5ad \
-    --image-in uns/spatial_pairwise_aligned/staining_image_transformed \
-    --output-mask uns/spatial_pairwise_aligned/mask_transformed_10px \
-    --model <path>/HE_cellpose_rajewsky \
-    --cellprob-threshold -1 \
-    --flow-threshold 2 \
-    --chunked \
-    --gpu \
-    --num-workers 32 \
-    --dilate-px 10
-
-### STEP 6 ###
-# assign the individual barcoded spots (0.5 micron) into the segmentation mask
-openst transcript_assign \
-    --adata <path>/<id>_stitched_spots_aligned.h5ad \
-    --spatial-key spatial_pairwise_aligned_fine \
-    --mask-in-adata \
-    --mask uns/spatial_pairwise_aligned/mask_transformed_10px \
-    --output <path>/<id>_segmented_cells_fine.h5ad
+# 7. create dge
+openst from_spacemake --project-id fc_sts_76 --sample-id fc_sts_76_3 transcript_assign --spatial-key obsm/spatial_pairwise_aligned_fine