Revision

examples/visualization/nglview_demo.ipynb

@@ -16,7 +16,7 @@
     {
      "data": {
       "application/vnd.jupyter.widget-view+json": {
-       "model_id": "4e25adfecfbc4102b206c79c6f171915",
+       "model_id": "318d91b0cf4f43c2b2216efa73a2d99f",
        "version_major": 2,
        "version_minor": 0
       },
@@ -29,7 +29,8 @@
    "source": [
     "import procaliper as pc\n",
     "import procaliper.view as pcv\n",
-    "from nglview.color import ColormakerRegistry"
+    "from nglview.color import ColormakerRegistry\n",
+    "import matplotlib as plt"
    ]
   },
   {
@@ -69,16 +70,16 @@
     }
    ],
    "source": [
-    "protein = pc.Protein.from_uniprot_id(\"A0A0B4J2F0\")\n",
+    "protein = pc.Protein.from_uniprot_id(\"P07900\")\n",
     "protein.fetch_pdb(save_path=\"scratch.pdb\")"
    ]
   },
   {
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "## Visualize SASA\n",
-    "First we compute the sasa data and assert that it was properly created. Then, we use the sasa values to create a color scheme and nglview widget using the `view` module of `procaliper`. We remove the default representation, register our created color scheme, and add the new representation with our color scheme."
+    "## Visualize pLDDT\n",
+    "First, we examine the pLDDT score for the AlphaFold structure we have fetched. We extract this using the `protein.get_confidence` method. These values fall between 0 and 100, and we rescale them to fall in the interval [0,1] for coloring. Since we have manually scaled these, we set `rescale=False`. Green and yellow regions indicate a lower confidence in local structure prediction. We can view the chosen color scale using `plt.colormaps.get_cmap(\"viridis_r\")`."
    ]
   },
   {
@@ -89,7 +90,82 @@
     {
      "data": {
       "application/vnd.jupyter.widget-view+json": {
-       "model_id": "b1fe72f9609f436d90e99a1f47932183",
+       "model_id": "d40e3b436a6f48a68bd9cd1fb17a235e",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "NGLWidget()"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "_ = protein.get_confidence()  # extract pLDDT values from the PDB\n",
+    "assert protein.confidence_data is not None\n",
+    "\n",
+    "pLDDT_scheme = pcv.ngl_scheme(\n",
+    "    [x / 100 for x in protein.confidence_data],\n",
+    "    color_mapper=\"viridis_r\",\n",
+    "    rescale=False,\n",
+    ")  # create a color scheme from the pLDDT values\n",
+    "\n",
+    "view = pcv.protein_to_nglview(protein)  # generate an nglview widget\n",
+    "view._remove_representation()  # remove the default representation\n",
+    "\n",
+    "cm.add_selection_scheme(\n",
+    "    \"all_pLDDT_value\", pLDDT_scheme\n",
+    ")  # add our color scheme to nglview\n",
+    "view.add_representation(\n",
+    "    \"ribbon\", color=\"all_pLDDT_value\"\n",
+    ")  # render the protein using our color scheme\n",
+    "\n",
+    "view"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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",
+      "text/html": [
+       "<div style=\"vertical-align: middle;\"><strong>viridis_r</strong> </div><div class=\"cmap\"><img alt=\"viridis_r colormap\" title=\"viridis_r\" style=\"border: 1px solid #555;\" src=\"data:image/png;base64,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\"></div><div style=\"vertical-align: middle; max-width: 514px; display: flex; justify-content: space-between;\"><div style=\"float: left;\"><div title=\"#fde725ff\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #fde725ff;\"></div> under</div><div style=\"margin: 0 auto; display: inline-block;\">bad <div title=\"#00000000\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #00000000;\"></div></div><div style=\"float: right;\">over <div title=\"#440154ff\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #440154ff;\"></div></div>"
+      ],
+      "text/plain": [
+       "<matplotlib.colors.ListedColormap at 0x204323a2660>"
+      ]
+     },
+     "execution_count": 5,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "plt.colormaps.get_cmap(\"viridis_r\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Visualize SASA\n",
+    "We compute the sasa data and assert that it was properly created. Then, we use the sasa values to create a color scheme and nglview widget using the `view` module of `procaliper`. We remove the default representation, register our created color scheme, and add the new representation with our color scheme."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "fa8aef4ce31541a58d5cdf6c3c65151a",
        "version_major": 2,
        "version_minor": 0
       },
@@ -132,13 +208,13 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 7,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "application/vnd.jupyter.widget-view+json": {
-       "model_id": "8188a2d779884abda606020394d14616",
+       "model_id": "57e8921c32824989bf0dc3249ce3bcbd",
        "version_major": 2,
        "version_minor": 0
       },
@@ -165,11 +241,18 @@
     "view.add_representation(\"surface\", color=\"all_charge_value\")\n",
     "view"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
   }
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": ".venv",
+   "display_name": "alphameter-py3.12",
    "language": "python",
    "name": "python3"
   },

procaliper/__init__.py

@@ -2,7 +2,7 @@
 
 __author__ = """AlphaMeter"""
 __email__ = "[email protected]"
-__version__ = "0.1.0"
+__version__ = "0.3.0"
 
 import procaliper.network as network
 import procaliper.protein_structure as protein_structure

procaliper/_protein.py

@@ -299,7 +299,13 @@ def get_biopython_residues(self) -> list[Residue]:
             raise ValueError("PDB location not set; use `fetch_pdb` first")
         p = PDBParser(QUIET=True)
         structure = p.get_structure("", self.pdb_location_absolute)
-        reslist = [res for model in structure for chain in model for res in chain]
+        reslist = [
+            res
+            for model in structure
+            for chain in model
+            for res in chain
+            if res.get_id()[0] == " "  # excludes heteroatoms and water
+        ]
         return reslist
 
     def get_confidence(self) -> list[float]:

procaliper/network.py

@@ -96,9 +96,7 @@ def regulatory_distance_network(protein: Protein) -> nx.Graph:
     all_regs = {**ptms, **binding, **active, **regions, **domains}
 
     # residues, excluding heteroatoms and water
-    protein_residues = [
-        res for res in protein.get_biopython_residues() if res.get_id()[0] == " "
-    ]
+    protein_residues = protein.get_biopython_residues()
 
     all_regs_residues = {}
     for k, v in all_regs.items():

procaliper/protein_structure/distance.py

@@ -91,7 +91,13 @@ def distance_matrix(
         npt.NDArray[np.float64]: distance matrix with shape nxn where n is the
             number of residues in the structure.
     """
-    residues = [res for model in structure for chain in model for res in chain]
+    residues = [
+        res
+        for model in structure
+        for chain in model
+        for res in chain
+        if res.get_id()[0] == " "
+    ]
     residues = list(enumerate(residues))
     adj = np.ones((len(residues), len(residues))) * np.inf
 

procaliper/view/nglview_utils.py

@@ -2,6 +2,8 @@
 
 from typing import Callable
 
+import matplotlib as plt
+
 from .._protein import Protein
 
 """
@@ -49,39 +51,58 @@ def _default_float_to_hex_rb(x: float) -> str:
         return f"#{int(1*255):02x}{int((1-x)*255):02x}{int((1-x)*255):02x}"
 
 
+def _matplotlib_cmapper(cmap: str, two_sided: bool) -> Callable[[float], str]:
+    if two_sided:
+        return lambda x: plt.colors.to_hex(plt.colormaps.get_cmap(cmap)(x / 2 + 0.5))
+
+    return lambda x: plt.colors.to_hex(plt.colormaps.get_cmap("viridis_r")(x))
+
+
 def ngl_scheme(
     data: list[float],
-    float_to_hex: Callable[[float], str] | None = None,
+    color_mapper: Callable[[float], str] | None = None,
     two_sided: bool = False,
+    rescale: bool = True,
 ) -> list[tuple[str, str]]:
     """Converts a list of values to an nglview color scheme.
 
     Args:
         data (list[float]): The list of values to convert.
-        float_to_hex (Callable[[float], str] | None, optional): Function that
-            converts a float to a hex color in the form `"#RRGGBB"`. If `None`,
-            a default function is used that interpolates between white and green
-            (one-sided) or red and blue (two-sided). Defaults to `None`.
+        color_mapper (Callable[[float], str] | str | None, optional): Function that
+            converts a float to a hex color in the form `"#RRGGBB"`. If a string, it
+            should be the name of a matplotlib colormap. If `None`, a default function
+            is used that interpolates between white and green (one-sided) or red and blue (two-sided).
+            Defaults to `None`.
         two_sided (bool, optional): Whether to use a two-sided color scheme. If
             `False`, we assume `data` only contains positive values. Defaults to
             `False`.
+        rescale (bool, optional): Whether to rescale the values to be between
+            0 and 1 (one-sided) or -1 and 1 (two-sided). Defaults to `True`.
+            If `False`, the values are not rescaled, but are assumed to fall within
+            [0, 1] or [-1, 1] depending on `two_sided`.
 
     Returns:
         list[tuple[str, str]]: A list of color and residue number tuples that
             are compatible with nglview.
     """
-    if float_to_hex is None:
+    if color_mapper is None:
         if two_sided:
-            float_to_hex = _default_float_to_hex_rb
+            color_mapper = _default_float_to_hex_rb
         else:
-            float_to_hex = _default_float_to_hex
+            color_mapper = _default_float_to_hex
+
+    if isinstance(color_mapper, str):
+        color_mapper = _matplotlib_cmapper(color_mapper, two_sided)
 
-    maxx = max(data)
-    scale = max(min(data), abs(maxx)) if two_sided else maxx
+    if rescale:
+        maxx = max(data)
+        scale = max(min(data), abs(maxx)) if two_sided else maxx
 
-    if scale == 0:
-        data_scaled = [0.0] * len(data)
+        if scale == 0:
+            data_scaled = [0.0] * len(data)
+        else:
+            data_scaled = [x / maxx for x in data]
     else:
-        data_scaled = [x / maxx for x in data]
+        data_scaled = data
 
-    return [(float_to_hex(x), f"{i+1}") for i, x in enumerate(data_scaled)]
+    return [(color_mapper(x), f"{i+1}") for i, x in enumerate(data_scaled)]

pyproject.toml

@@ -1,7 +1,7 @@
 [tool]
 [tool.poetry]
 name = "procaliper"
-version = "0.2.1"
+version = "0.3.0"
 homepage = "https://github.com/LifeWorks/procaliper"
 description = "Skeleton project created by Python Project Wizard (ppw)."
 authors = ["AlphaMeter <[email protected]>"]