read_tenx_visium executes successfully

Author: keviny2

src/spatialexperiment/__init__.py

@@ -16,6 +16,7 @@
     del version, PackageNotFoundError
 
 from .ProxySFE import ProxySpatialFeatureExperiment
+from .io import read_tenx_visium
 from .SpatialExperiment import SpatialExperiment
 from .SpatialImage import (
     LoadedSpatialImage,
@@ -24,3 +25,15 @@
     VirtualSpatialImage,
     construct_spatial_image_class,
 )
+
+__all__ = [
+    "ProxySpatialFeatureExperiment",
+    "read_tenx_visium",
+    "SpatialExperiment",
+    "LoadedSpatialImage",
+    "RemoteSpatialImage",
+    "StoredSpatialImage",
+    "VirtualSpatialImage",
+    "construct_spatial_image_class",
+]
+

src/spatialexperiment/_imgutils.py

@@ -1,6 +1,9 @@
 from typing import Union
 
 import os
+from io import BytesIO
+import requests
+from PIL import Image
 from biocframe import BiocFrame
 from .SpatialImage import construct_spatial_image_class
 
@@ -9,24 +12,65 @@
 __license__ = "MIT"
 
 
+
+def read_image(input_image):
+    """Read image from PIL Image, file path, or URL.
+
+    Args:
+        input_image: PIL Image, string path to local file, or URL string.
+
+    Returns:
+        The loaded image.
+
+    Raises:
+        TypeError: If input is not PIL Image, path string, or URL string.
+    """
+    if isinstance(input_image, Image.Image):
+        return input_image
+    
+    if isinstance(input_image, str):
+        if input_image.startswith(('http://', 'https://')):
+            response = requests.get(input_image)
+            return Image.open(BytesIO(response.content))
+        else:
+            return Image.open(input_image)
+            
+    raise TypeError(f"Expected PIL Image, path, or URL. Got {type(input_image)}")
+
+
 def get_img_data(
     img: Union[str, os.PathLike],
     scale_factor: str,
     sample_id: str,
     image_id: str,
+    load: bool = True
 ) -> BiocFrame:
     """
     Construct an image data dataframe.
 
     Args:
-        img: A path or url to the image file.
-        scale_factor (str): The scale factor associated with the image.
-        sample_id (str): A unique identifier for the sample to which the image belongs.
-        image_id (str): A unique identifier for the image itself.
+        img:
+            A path or url to the image file.
+
+        scale_factor:
+            The scale factor associated with the image.
+
+        sample_id:
+            A unique identifier for the sample to which the image belongs.
+
+        image_id:
+            A unique identifier for the image itself.
+
+        load:
+            A boolean specifying whether the image(s) should be loaded into memory? If False, will store the path/URL instead.
+            Defaults to `True`.
 
     Returns:
         The image data.
     """
+    if load:
+        img = read_image(img)
+
     spi = construct_spatial_image_class(img)
     return BiocFrame(
         {
@@ -47,7 +91,9 @@ def retrieve_rows_by_id(
     Retrieve rows from `img_data` based on specified `sample_id` and `image_id`.
 
     Args:
-        img_data: The data from which to retrieve rows.
+        img_data:
+            The data from which to retrieve rows.
+
         sample_id:
             - `sample_id=True`: Matches all samples.
             - `sample_id=None`: Matches the first sample.

src/spatialexperiment/_initutils.py

@@ -3,7 +3,7 @@
 
 from biocframe import BiocFrame
 from PIL import Image
-from SpatialImage import SpatialImage
+from .SpatialImage import construct_spatial_image_class
 from summarizedexperiment._frameutils import _sanitize_frame
 
 __author__ = "keviny2"
@@ -96,7 +96,7 @@ def construct_img_data(
     spis = []
     for image_source in image_sources:
         result = Image.open(image_source) if load_image else image_source
-        spi = SpatialImage(result)
+        spi = construct_spatial_image_class(result)
         spis.append(spi)
 
     img_data = {

src/spatialexperiment/io/__init__.py

@@ -0,0 +1,4 @@
+from .tenx_visium import read_tenx_visium
+
+__all__ = ["read_tenx_visium"]
+

src/spatialexperiment/read_10x_visium.py src/spatialexperiment/io/tenx_visium.pysrc/spatialexperiment/read_10x_visium.py renamed to src/spatialexperiment/io/tenx_visium.py

@@ -7,14 +7,14 @@
 import pandas as pd
 
 from biocframe import BiocFrame
-from biocutils import is_list_of_type
+import biocutils as ut
 from singlecellexperiment import read_tenx_mtx
-from spatialexperiment import SpatialExperiment
-from ._imgutils import get_img_data
-from ._initutils import construct_spatial_coords_from_names
+from ..SpatialExperiment import SpatialExperiment
+from .._imgutils import get_img_data
+from .._initutils import construct_spatial_coords_from_names
 
 
-def read_tissue_positions(tissue_positions_paths):
+def read_tissue_positions(tissue_positions_path):
     column_names = [
         "barcode",
         "in_tissue",
@@ -24,30 +24,23 @@ def read_tissue_positions(tissue_positions_paths):
         "pxl_col_in_fullres",
     ]
 
-    tissue_positions_combined = []
-    for i, tissue_positions_path in enumerate(tissue_positions_paths):
-        has_header = "list" not in os.path.basename(tissue_positions_path)
+    has_header = "list" not in os.path.basename(tissue_positions_path)
 
-        tissue_positions = pd.read_csv(
-            tissue_positions_path, header=0 if has_header else None, index_col=0
-        )
-        tissue_positions.columns = column_names
-
-        if len(tissue_positions_paths) > 1:
-            tissue_positions.index = [f"{i+1}_{idx}" for idx in tissue_positions.index]
-
-        tissue_positions["in_tissue"] = tissue_positions["in_tissue"].astype(bool)
-        tissue_positions_combined.append(tissue_positions)
+    tissue_positions = pd.read_csv(
+        tissue_positions_path, header=0 if has_header else None, names=column_names
+    )
+    tissue_positions = tissue_positions.set_index("barcode")
+    tissue_positions["in_tissue"] = tissue_positions["in_tissue"].astype(bool)
 
-    tissue_positions_combined = pd.concat(tissue_positions_combined)
-    return tissue_positions_combined
+    return tissue_positions
 
 
 def read_img_data(
     path: str = ".",
     sample_ids: Optional[List[str]] = None,
     image_sources: Optional[List[str]] = None,
     scale_factors: str = None,
+    load: bool = True
 ) -> BiocFrame:
     """Read in images and scale factors for 10x Genomics Visium data, and return as a valid `img_data` object.
 
@@ -63,6 +56,10 @@ def read_img_data(
 
         scale_factors:
             The .json file where to find the scale factors.
+
+        load:
+            A boolean specifying whether the image(s) should be loaded into memory? If False, will store the path/URL instead.
+            Defaults to `True`.
     """
     # get sample identifiers
     if sample_ids is None:
@@ -120,17 +117,18 @@ def read_img_data(
                 scale_factor=scale_factor,
                 sample_id=sample_id,
                 image_id=image_id,
+                load=load
             )
             img_data = img_data.combine_rows(curr_image_data)
 
     return img_data
 
 
-def read_10x_visium(
+def read_tenx_visium(
     samples: List[Union[str, os.PathLike]],
     sample_ids: Optional[List[str]] = None,
     type: str = "HDF5",
-    data: str = ["filtered", "raw"],
+    data: str = "filtered",
     images: List[str] = "lowres",
     load: bool = True,
 ):
@@ -154,6 +152,7 @@ def read_10x_visium(
 
         load:
             A boolean specifying whether the image(s) should be loaded into memory? If False, will store the path/URL instead.
+            Defaults to `True`.
     """
     # check validity of input arguments
     allowed_types = ["HDF5", "sparse", "auto", "prefix"]
@@ -177,7 +176,7 @@ def read_10x_visium(
 
     if sample_ids is None:
         sample_ids = [f"sample{str(i).zfill(2)}" for i in range(1, len(samples) + 1)]
-    elif not is_list_of_type(sample_ids, str) and len(set(sample_ids)) != len(samples):
+    elif not ut.is_list_of_type(sample_ids, str) and len(set(sample_ids)) != len(samples):
         raise ValueError(
             "`sample_ids` should contain as many unique values as `samples`."
         )
@@ -261,13 +260,18 @@ def read_10x_visium(
         sample_ids=sample_ids,
         image_sources=image_file_paths,
         scale_factors=scale_factors_paths,
+        load=load
     )
 
+    spes = []
     for i, counts_dir_path in enumerate(counts_dir_paths):
         sce = read_tenx_mtx(counts_dir_path)
         tissue_positions = read_tissue_positions(tissue_positions_paths[i])
 
-        obs = set(sce.col_names).intersection(set(tissue_positions.index))
+        barcodes = sce.column_data["barcode"]
+        sce = sce.set_column_names(barcodes)
+
+        obs = list(set(sce.col_names).intersection(set(tissue_positions.index)))
         sce = sce[:, obs]
 
         tissue_positions = tissue_positions.loc[obs, :]
@@ -279,9 +283,17 @@ def read_10x_visium(
 
         spe = SpatialExperiment(
             assays=sce.assays,
-            row_data=sce.row_data["Symbol"],
+            row_data=BiocFrame(
+                {
+                    "symbol": sce.row_data["gene_symbols"]
+                }
+            ),
             column_data=column_data,
             spatial_coords=spatial_coords
         )
+        spes.append(spe)
+
+    spe_combined = ut.combine_columns(*spes)
+    spe_combined.img_data = image
 
-        # TODO: implement combine ops on SpatialExperiment objects
+    return spe_combined

tests/10xVisium/section1/outs/raw_feature_bc_matrix/barcodes.tsv

@@ -0,0 +1,50 @@
+AAACAACGAATAGTTC-1
+AAACAAGTATCTCCCA-1
+AAACAATCTACTAGCA-1
+AAACACCAATAACTGC-1
+AAACAGAGCGACTCCT-1
+AAACAGCTTTCAGAAG-1
+AAACAGGGTCTATATT-1
+AAACAGTGTTCCTGGG-1
+AAACATGGTGAGAGGA-1
+AAACATTTCCCGGATT-1
+AAACCACTACACAGAT-1
+AAACCCGAACGAAATC-1
+AAACCGGAAATGTTAA-1
+AAACCGGGTAGGTACC-1
+AAACCGTTCGTCCAGG-1
+AAACCTAAGCAGCCGG-1
+AAACCTCATGAAGTTG-1
+AAACGAAGAACATACC-1
+AAACGAAGATGGAGTA-1
+AAACGACAGTCTTGCC-1
+AAACGAGACGGTTGAT-1
+AAACGCCCGAGATCGG-1
+AAACGCTGGGCACGAC-1
+AAACGGGCGTACGGGT-1
+AAACGGGTTGGTATCC-1
+AAACGGTTGCGAACTG-1
+AAACGTGTTCGCCCTA-1
+AAACTAACGTGGCGAC-1
+AAACTCGGTTCGCAAT-1
+AAACTCGTGATATAAG-1
+AAACTGCTGGCTCCAA-1
+AAACTTAATTGCACGC-1
+AAACTTGCAAACGTAT-1
+AAAGAATGACCTTAGA-1
+AAAGAATGTGGACTAA-1
+AAAGACATGAAGTTTA-1
+AAAGACCCAAGTCGCG-1
+AAAGACTGGGCGCTTT-1
+AAAGCTTGCCTACATA-1
+AAAGGCCCTATAATAC-1
+AAAGGCTACGGACCAT-1
+AAAGGCTCTCGCGCCG-1
+AAAGGGATGTAGCAAG-1
+AAAGGGCAGCTTGAAT-1
+AAAGGTAAGCTGTACC-1
+AAAGGTCAACGACATG-1
+AAAGTAGCATTGCTCA-1
+AAAGTCACTGATGTAA-1
+AAAGTCGACCCTCAGT-1
+AAAGTGCCATCAATTA-1

tests/10xVisium/section1/outs/raw_feature_bc_matrix/features.tsv

@@ -0,0 +1,50 @@
+ENSMUSG00000051951	Xkr4	Gene Expression
+ENSMUSG00000089699	Gm1992	Gene Expression
+ENSMUSG00000102343	Gm37381	Gene Expression
+ENSMUSG00000025900	Rp1	Gene Expression
+ENSMUSG00000025902	Sox17	Gene Expression
+ENSMUSG00000104328	Gm37323	Gene Expression
+ENSMUSG00000033845	Mrpl15	Gene Expression
+ENSMUSG00000025903	Lypla1	Gene Expression
+ENSMUSG00000104217	Gm37988	Gene Expression
+ENSMUSG00000033813	Tcea1	Gene Expression
+ENSMUSG00000002459	Rgs20	Gene Expression
+ENSMUSG00000085623	Gm16041	Gene Expression
+ENSMUSG00000033793	Atp6v1h	Gene Expression
+ENSMUSG00000025905	Oprk1	Gene Expression
+ENSMUSG00000033774	Npbwr1	Gene Expression
+ENSMUSG00000025907	Rb1cc1	Gene Expression
+ENSMUSG00000090031	4732440D04Rik	Gene Expression
+ENSMUSG00000087247	Alkal1	Gene Expression
+ENSMUSG00000033740	St18	Gene Expression
+ENSMUSG00000051285	Pcmtd1	Gene Expression
+ENSMUSG00000097797	Gm26901	Gene Expression
+ENSMUSG00000103067	Gm30414	Gene Expression
+ENSMUSG00000025909	Sntg1	Gene Expression
+ENSMUSG00000061024	Rrs1	Gene Expression
+ENSMUSG00000025911	Adhfe1	Gene Expression
+ENSMUSG00000067879	3110035E14Rik	Gene Expression
+ENSMUSG00000099827	Gm29520	Gene Expression
+ENSMUSG00000025912	Mybl1	Gene Expression
+ENSMUSG00000045210	Vcpip1	Gene Expression
+ENSMUSG00000097893	1700034P13Rik	Gene Expression
+ENSMUSG00000025915	Sgk3	Gene Expression
+ENSMUSG00000046101	Mcmdc2	Gene Expression
+ENSMUSG00000098234	Snhg6	Gene Expression
+ENSMUSG00000099032	Tcf24	Gene Expression
+ENSMUSG00000025916	Ppp1r42	Gene Expression
+ENSMUSG00000087199	Gm15818	Gene Expression
+ENSMUSG00000025917	Cops5	Gene Expression
+ENSMUSG00000056763	Cspp1	Gene Expression
+ENSMUSG00000067851	Arfgef1	Gene Expression
+ENSMUSG00000042501	Cpa6	Gene Expression
+ENSMUSG00000048960	Prex2	Gene Expression
+ENSMUSG00000057715	A830018L16Rik	Gene Expression
+ENSMUSG00000097171	Gm17644	Gene Expression
+ENSMUSG00000101314	Gm29663	Gene Expression
+ENSMUSG00000016918	Sulf1	Gene Expression
+ENSMUSG00000025938	Slco5a1	Gene Expression
+ENSMUSG00000099498	Gm29283	Gene Expression
+ENSMUSG00000042414	Prdm14	Gene Expression
+ENSMUSG00000005886	Ncoa2	Gene Expression
+ENSMUSG00000101476	Gm29570	Gene Expression