update README with runnable test data

Author: trangle1302

README.md

@@ -40,16 +40,17 @@ Before running the analysis, you need to:
 
 - Follow the workflow steps, which are described in detail in the manuscript and summarized below in the [Shapespace construction](## Shapespace construction) section.
 
-We also provide a test dataset (single-cell crops and masks) which allows you to quickly test shape parameterization, constructing shapespace and map protein intensity to the average cell shape.
+We also provide a test dataset (single-cell crops and masks) which allows you to quickly test shape parameterization, constructing shapespace and map protein intensity to the average cell shape. NOTE: This dataset is intentionally small for testing and may not preserve the true average shape representation of the cell line. The number of samples/organelles is also limited, so it will not recover the true organelle map.
 
 ```bash
 wget https://ell-vault.stanford.edu/dav/trangle/www/K-562.zip
 unzip K-562.zip -d K-562
 
 python -m coefficients.s2_calculate_fft
+python -m analysis.cell_nucleus_ratio
 python -m shapemodes.s3_calculate_shapemodes
-python -m warp.s4_concentric_rings_intensity # check cfg.N_ISOS and cfg.LANDMARKS
-python -m warp.s4_protein_image_warp # check cfg.LANDMARKS
+python -m warps.s4_concentric_rings_intensity --cell_line K-562  --n_isos 10 20 # check cfg.N_ISOS and cfg.LANDMARKS
+python -m warps.s4_tsp --cell_line K-562 # check cfg.LANDMARKS
 ```
 For large datasets or when analyzing multiple cell lines, consider using a workflow manager such as **Snakemake**, or submitting separate jobs to a compute cluster using **SLURM**. Example workflow files and job scripts can be found inside each folder. 
 

configs/config.py

@@ -109,5 +109,5 @@
     PROJECT_DIR = f"/scratch/groups/emmalu/2Dshapespace/{CELL_LINE.replace(' ','_')}"
     META_PATH = "/scratch/groups/emmalu/sl_pHPA_15_0.05_euclidean_100000_rmoutliers_ilsc_3d_bbox_rm_border.csv"
 else:
-    PROJECT_DIR = f"./{CELL_LINE.replace(' ','_')}"
-    META_PATH = "" # your own metadata to perform downstream analysis
+    PROJECT_DIR = f"./{CELL_LINE}"
+    META_PATH = f"./{CELL_LINE}/meta_k562.csv" # your own metadata to perform downstream analysis

setup.py

@@ -0,0 +1,22 @@
+from setuptools import setup, find_packages
+
+setup(
+    name="2D_shapespace",  # Package name
+    version="0.1.0",
+    description="Analysis of cell line shapespace and organelle correlations",
+    author="trangle1302",
+    packages=find_packages(),
+    python_requires=">=3.8",
+    install_requires=[
+        "scikit-learn==1.4.2",
+        "scikit-image==0.22.0",
+        "seaborn==0.13.0",
+        "opencv-python-headless==4.6.0.66",
+        "more_itertools==10.7.0",
+        "tqdm==4.67.1",
+        "numpy==1.26.4",
+        # Optional / experimental (not used in final version):
+        # "pyefd",
+        # "PyWavelets",
+    ],
+)

warps/s4_concentric_rings_intensity.py

@@ -111,8 +111,14 @@ def main():
         mappings = pd.read_csv('/data/HPA-IF-images/IF-image.csv')
         mappings = mappings[mappings.atlas_name=='U2OS']
     else:
-        cellline_meta = os.path.join(project_dir, os.path.basename(cfg.META_PATH).replace(".csv", "_splitVesiclesPCP.csv"))
-        mappings = pd.read_csv(cellline_meta)
+        cellline_meta_path = os.path.join(project_dir, os.path.basename(cfg.META_PATH).replace(".csv", "_splitVesiclesPCP.csv"))
+        if os.path.exists(cellline_meta_path):
+            mappings = pd.read_csv(cellline_meta_path)
+        else:
+            mappings = pd.read_csv(cfg.META_PATH)
+            mappings["cell_idx"] = [idx.split("_", 1)[1] for idx in mappings.id]
+            from warps.avg_organelle import unmerge_label
+            mappings = unmerge_label(mappings)
     #mappings = mappings[~mappings.sc_target.isin(["Negative","Multi-Location"])]
     print(mappings.columns)
     #print(mappings.sc_target.value_counts(), mappings.cell_idx)

warps/s4_tsp.py

@@ -200,8 +200,14 @@ def main():
     if not os.path.exists(protein_dir):
         os.makedirs(protein_dir)
 
-    cellline_meta = os.path.join(project_dir, os.path.basename(cfg.META_PATH).replace(".csv", "_splitVesiclesPCP.csv"))
-    mappings = pd.read_csv(cellline_meta)
+    cellline_meta_path = os.path.join(project_dir, os.path.basename(cfg.META_PATH).replace(".csv", "_splitVesiclesPCP.csv"))
+    if os.path.exists(cellline_meta_path):
+        mappings = pd.read_csv(cellline_meta_path)
+    else:
+        mappings = pd.read_csv(cfg.META_PATH)
+        mappings["cell_idx"] = [idx.split("_", 1)[1] for idx in mappings.id]
+        from warps.avg_organelle import unmerge_label
+        mappings = unmerge_label(mappings)
     log_dir = f"{project_dir}/logs"
     fft_dir = f"{project_dir}/fftcoefs/{cfg.ALIGNMENT}"
     fft_path = os.path.join(fft_dir, f"fftcoefs_{cfg.N_COEFS}.txt")