data_barseq2_mouse_viscortex

data/barseq2_mouse_viscortex/viscortex.py

@@ -0,0 +1,135 @@
+#!/usr/bin/env python
+
+# Made by Paul Kiessling [email protected]
+
+import argparse
+import json
+import os
+import tempfile
+import urllib.request
+
+import anndata
+import pandas as pd
+import scipy
+
+LINKS = {
+    "slice_1": "https://gene.ai.tencent.com/SpatialOmics/api/download_data/2492",
+    "slice_2": "https://gene.ai.tencent.com/SpatialOmics/api/download_data/2494",
+    "slice_3": "https://gene.ai.tencent.com/SpatialOmics/api/download_data/2493",
+}
+
+META_DICT = {"technology": "BARseq2"}
+LICENSE = """CC BY 4.0"""
+SAMPLE_COLUMNS = [
+    "patient",
+    "sample",
+    "position",
+    "replicate",
+    "n_clusters",
+    "directory",
+]
+
+
+def download_links(links, temp_dir):
+    for name, link in links.items():
+        try:
+            response = urllib.request.urlopen(link)
+
+            filename = os.path.join(temp_dir, name + ".h5ad")
+            with open(filename, "wb") as file:
+                file.write(response.read())
+            print(f"Downloaded: {filename}")
+        except Exception as e:
+            print(f"Error downloading {link}: {e}")
+
+
+def process_adata(adata_path, output_folder, iteration, sample_df):
+    folder_name = os.path.splitext(os.path.basename(adata_path))[0]
+    complete_path = os.path.join(output_folder, folder_name)
+    os.makedirs(complete_path, exist_ok=True)
+    adata = anndata.read_h5ad(adata_path)
+
+    # Observations
+    obs = adata.obs.copy()
+    obs["selected"] = "true"
+    obs.loc[obs.layer == "outside_VISp", "selected"] = "false"
+    obs.to_csv(f"{complete_path}/observations.tsv", sep="\t", index_label="")
+
+    # Features
+    vars = adata.var.copy()
+    vars["selected"] = "true"
+    vars.to_csv(f"{complete_path}/features.tsv", sep="\t", index_label="")
+
+    # Coordinates
+    coords = pd.DataFrame(adata.obsm["spatial"], columns=["x", "y"])
+    coords.index = adata.obs.index
+    coords.to_csv(f"{complete_path}/coordinates.tsv", sep="\t", index_label="")
+
+    # Matrix
+    scipy.io.mmwrite(f"{complete_path}/counts.mtx", adata.X)
+
+    # add info for sample.tsv
+    # Your sample_data_basis dictionary
+    sample_data_basis = {
+        "patient": "0",
+        "sample": iteration,
+        "position": "0",
+        "replicate": "0",
+        "n_clusters": adata.obs.layer.nunique()
+        - 1,  # -1 for outside_VISp which we exclude
+        "directory": folder_name,
+    }
+
+    # Concatenating the new DataFrame to sample_df
+    sample_df.iloc[iteration] = sample_data_basis
+    # Write labels.tsv
+    labels = adata.obs.loc[adata.obs.layer != "outside_VISp", "layer"]
+    labels.to_csv(f"{complete_path}/labels.tsv", sep="\t", index_label="")
+
+
+def write_json(dict, output_path):
+    with open(output_path, "w") as json_file:
+        json.dump(dict, json_file)
+
+
+def main():
+    # Set up command-line argument parser
+    parser = argparse.ArgumentParser(
+        description="Convert Barseq2 Viscortex data to Spacehack format."
+    )
+
+    # Add arguments for output folder
+    parser.add_argument(
+        "-o", "--out_dir", help="Output directory to write files to.", required=True
+    )
+
+    # Parse the command-line arguments
+    args = parser.parse_args()
+
+    # Download and process
+    with tempfile.TemporaryDirectory() as temp_dir:
+        download_links(LINKS, temp_dir)
+        os.makedirs(args.out_dir, exist_ok=True)
+
+        sample_df = pd.DataFrame(columns=SAMPLE_COLUMNS, index=range(len(LINKS)))
+        anndatas = [
+            os.path.join(temp_dir, file)
+            for file in os.listdir(temp_dir)
+            if file.endswith(".h5ad")
+        ]
+        for iteration, adata in enumerate(anndatas):
+            process_adata(adata, args.out_dir, iteration, sample_df)
+
+        # write json
+        write_json(META_DICT, f"{args.out_dir}/experiment.json")
+
+        # write samples.tsv
+        sample_df.to_csv(f"{args.out_dir}/samples.tsv", sep="\t", index_label=False)
+
+        # write LICENSE
+        with open(f"{args.out_dir}/LICENSE.md", "w") as file:
+            file.write(LICENSE)
+
+
+if __name__ == "__main__":
+    main()

data/barseq2_mouse_viscortex/viscortex.yml

@@ -0,0 +1,8 @@
+channels:
+  - conda-forge
+dependencies:
+  - python=3.11.6
+  - scipy=1.11.4
+  - anndata=0.10.3
+  - numpy=1.26.2
+  - pandas=2.1.3