spatial plotting

Author: pcahan1

src/pySingleCellNet/plotting/dot.py

@@ -19,9 +19,109 @@
 from palettable.scientific.sequential import Batlow_20
 from anndata import AnnData
 from scipy.sparse import csr_matrix
-from sklearn.metrics import f1_score
+from scipy import sparse
+from typing import Optional, Callable
 from ..utils import *
 
+from scipy import sparse
+from typing import Optional, Callable
+
+def spatial_two_genes(
+    adata: AnnData,
+    gene1: str,
+    gene2: str,
+    title: Optional[str] = None,
+    scale_max_value: float = 2.0,
+    spot_size: float = 35,
+    cmap: str = 'RdBu_r',
+    alpha: float = 0.5,
+    copy_adata: bool = True,
+    combine_fun: Optional[Callable[[np.ndarray, np.ndarray], np.ndarray]] = None,
+    **plot_kwargs
+) -> None:
+    """Plot a custom combination of two gene expressions on a spatial scatter,
+    scaling only those two genes.
+
+    This will (optionally) make a copy of your AnnData, extract expression for
+    gene1 and gene2, scale each gene vector (zero‐center, unit variance,
+    clipped to ±scale_max_value), compute a per‐cell combined metric, store it
+    in `.obs[title]`, and then call `sc.pl.spatial`.
+
+    Args:
+        adata: AnnData with spatial coords and expression in `.X` (or `.layers`).
+        gene1: Name of the first gene (must be in `adata.var_names`).
+        gene2: Name of the second gene.
+        title: Key under which to store the combined metric in `adata.obs` and
+            plot title. Defaults to `"gene1_gene2"`.
+        scale_max_value: Maximum absolute value to clip the scaled gene vectors.
+            Defaults to 2.0.
+        spot_size: Passed to `sc.pl.spatial(..., spot_size=...)`. Default 35.
+        cmap: Colormap for `sc.pl.spatial`. Default `'RdBu_r'`.
+        alpha: Spot transparency for `sc.pl.spatial`. Default 0.5.
+        copy_adata: If True, operate on a copy. Otherwise overwrite `adata.obs`.
+        combine_fun: Function `(g1, g2) -> combined`. If None, uses
+            `(g1 * g2) + g1 - g2`.
+        **plot_kwargs: Any additional args forwarded to `sc.pl.spatial`.
+
+    Returns:
+        None. Displays a spatial scatter of the combined metric.
+    """
+    # determine obs key / title
+    if title is None:
+        title = f"{gene1}_{gene2}"
+
+    # copy or in-place
+    ad = adata.copy() if copy_adata else adata
+
+    # extract raw expression
+    X1 = ad[:, gene1].X
+    X2 = ad[:, gene2].X
+
+    # to 1D numpy arrays
+    def to_array(mat):
+        if sparse.issparse(mat):
+            arr = mat.A.flatten()
+        else:
+            arr = np.asarray(mat).flatten()
+        return arr
+
+    g1 = to_array(X1)
+    g2 = to_array(X2)
+
+    # scale each gene individually
+    def scale_vec(x):
+        m = x.mean()
+        s = x.std(ddof=0)
+        if s == 0:
+            # avoid divide by zero
+            scaled = x - m
+        else:
+            scaled = (x - m) / s
+        return np.clip(scaled, -scale_max_value, scale_max_value)
+
+    g1_scaled = scale_vec(g1)
+    g2_scaled = scale_vec(g2)
+
+    # combine
+    if combine_fun is None:
+        expr = (g1_scaled * g2_scaled) + g1_scaled - g2_scaled
+    else:
+        expr = combine_fun(g1_scaled, g2_scaled)
+
+    # store and plot
+    ad.obs[title] = pd.Series(expr, index=ad.obs.index)
+    sc.pl.spatial(
+        ad,
+        color=title,
+        spot_size=spot_size,
+        cmap=cmap,
+        alpha=alpha,
+        **plot_kwargs
+    )
+
+
+
+
 def umi_counts_ranked(adata, total_counts_column="total_counts"):
     """
     Identifies and plors the knee point of the UMI count distribution in an AnnData object.

src/pySingleCellNet/plotting/scatter.py

@@ -73,41 +73,3 @@ def scatter_qc_adata(adata, title_suffix=""):
     plt.show()
 
 
-
-
-
-
-
-
-
-
-def scatter_qc_adata(adata, title_suffix=""):
-    # Extract necessary columns from the adata object
-    total_counts = adata.obs['total_counts']
-    n_genes_by_counts = adata.obs['n_genes_by_counts']
-    pct_counts_mt = adata.obs['pct_counts_mt']
-
-    # Create a figure with two subplots (1 row, 2 columns)
-    fig, axes = plt.subplots(1, 2, figsize=(12, 6))
-
-    # First subplot: total_counts vs n_genes_by_counts, colored by pct_counts_mt
-    scatter1 = axes[0].scatter(total_counts, n_genes_by_counts, c=pct_counts_mt, cmap='viridis', alpha=0.5, s=1)
-    axes[0].set_xlabel(f'Total Counts ({title_suffix})')
-    axes[0].set_ylabel(f'Number of Genes by Counts ({title_suffix})')
-    axes[0].set_title(f'Total Counts vs Genes ({title_suffix})')
-    # Add a colorbar
-    fig.colorbar(scatter1, ax=axes[0], label='% Mito')
-
-    # Second subplot: n_genes_by_counts vs pct_counts_mt
-    scatter2 = axes[1].scatter(n_genes_by_counts, pct_counts_mt, alpha=0.5, s=1)
-    axes[1].set_xlabel(f'Number of Genes by Counts ({title_suffix})')
-    axes[1].set_ylabel(f'% Mito ({title_suffix})')
-    axes[1].set_title(f'Genes vs % Mito ({title_suffix})')
-
-    # Adjust layout to avoid overlap
-    plt.tight_layout()
-
-    # Show the plot
-    plt.show()
-
-

src/pySingleCellNet/utils/gene.py

@@ -92,7 +92,8 @@ def find_knn_modules(
     n_pcs = elbow + npcs_adjust
 
     # Compute the kNN graph using correlation as the metric
-    sc.pp.neighbors(adata_transposed, n_neighbors=knn, n_pcs=n_pcs, metric='correlation')
+    # sc.pp.neighbors(adata_transposed, n_neighbors=knn, n_pcs=n_pcs, metric='correlation')
+    sc.pp.neighbors(adata_transposed, n_neighbors=knn, n_pcs=n_pcs, metric='euclidean')
 
     # Perform Leiden clustering with an explicit resolution keyword
     sc.tl.leiden(adata_transposed, resolution=leiden_resolution)