Added check to keep consistency with alphabet (#143)

* Added check to keep consistency with alphabet

* Fixed pre-commit maybe

* Now we play find the correct langchain version

.pre-commit-config.yaml

@@ -7,13 +7,11 @@ repos:
       - id: end-of-file-fixer
       - id: mixed-line-ending
   - repo: https://github.com/psf/black
-    rev: "22.3.0"
+    rev: "23.11.0"
     hooks:
-    - id: black
-  - repo: https://github.com/tomcatling/black-nb
-    rev: "0.7"
-    hooks:
-      - id: black-nb
-        description: strip output and black source
+      - id: black
         additional_dependencies: ['black[jupyter]']
-        args: ["--clear-output"]
+  - repo: https://github.com/kynan/nbstripout
+    rev: "0.6.0"
+    hooks:
+      - id: nbstripout

exmol/exmol.py

@@ -1,4 +1,4 @@
-from functools import reduce
+from functools import reduce, lru_cache
 import inspect
 from typing import *
 import io
@@ -9,6 +9,7 @@
 from matplotlib.patches import Rectangle, FancyBboxPatch  # type: ignore
 from matplotlib.offsetbox import AnnotationBbox  # type: ignore
 import matplotlib as mpl  # type: ignore
+import re
 import selfies as sf  # type: ignore
 import tqdm  # type: ignore
 import textwrap  # type: ignore
@@ -367,6 +368,38 @@ def get_basic_alphabet() -> Set[str]:
     return a
 
 
+# The code below checks for accidental addition of symbols outside of the alphabet.
+# the only way this can happen is if the character indicating a ring
+# length is mutated to appear eleswhere. The ring length symbol
+# is always a plain uncharged element.
+
+
+def _alphabet_to_elements(alphabet: List[str]) -> Set[str]:
+    """Converts SELFIES alphabet to element symbols"""
+    symbols = []
+    for s in alphabet:
+        s = s.replace("[", "").replace("]", "")
+        if s.isalpha():
+            symbols.append(s)
+    return set(symbols)
+
+
+def _check_alphabet_consistency(
+    smiles: str, alphabet_symbols: Set[str], check=False
+) -> True:
+    """Checks if SMILES only contains tokens from alphabet"""
+
+    alphabet_symbols = _alphabet_to_elements(set(alphabet_symbols))
+    # find all elements in smiles (Upper alpha or upper alpha followed by lower alpha)
+    smiles_symbols = set(re.findall(r"[A-Z][a-z]?", smiles))
+    if check and not smiles_symbols.issubset(alphabet_symbols):
+        # show which symbols are not in alphabet
+        raise ValueError(
+            "symbols not in alphabet" + smiles_symbols.difference(alphabet_symbols)
+        )
+    return smiles_symbols.issubset(alphabet_symbols)
+
+
 def run_stoned(
     start_smiles: str,
     fp_type: str = "ECFP4",
@@ -392,6 +425,9 @@ def run_stoned(
         alphabet = get_basic_alphabet()
     if type(alphabet) == set:
         alphabet = list(alphabet)
+    alphabet_symbols = _alphabet_to_elements(alphabet)
+    # make sure starting smiles is consistent with alphabet
+    _ = _check_alphabet_consistency(start_smiles, alphabet_symbols, check=True)
     num_mutation_ls = list(range(min_mutations, max_mutations + 1))
 
     start_mol = smi2mol(start_smiles)
@@ -418,6 +454,16 @@ def run_stoned(
         )
         # Convert back to SMILES:
         smiles_back = [sf.decoder(x) for x in selfies_mut]
+        # check if smiles are consistent with alphabet and downslect
+        selfies_mut, smiles_back = zip(
+            *[
+                (s, sm)
+                for s, sm in zip(selfies_mut, smiles_back)
+                if _check_alphabet_consistency(sm, alphabet_symbols)
+            ]
+        )
+        selfies_mut, smiles_back = list(selfies_mut), list(smiles_back)
+
         all_smiles_collect = all_smiles_collect + smiles_back
         all_selfies_collect = all_selfies_collect + selfies_mut
         if _pbar:

exmol/stoned/stoned.py

@@ -446,7 +446,6 @@ def get_mutated_SELFIES(selfies_ls, num_mutations, alphabet):
     for _ in range(num_mutations):
         selfie_ls_mut_ls = []
         for str_ in selfies_ls:
-
             str_chars = get_selfie_chars(str_)
             max_molecules_len = len(str_chars) + num_mutations
 

exmol/version.py

@@ -1 +1 @@
-__version__ = "3.0.3"
+__version__ = "3.0.4"

paper2_LIME/RF-lime.ipynb

@@ -63,9 +63,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "scrolled": true
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "# make object that can compute descriptors\n",
@@ -249,9 +247,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "scrolled": false
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "x = wls_attr.keys()\n",

paper2_LIME/Solubility-RNN.ipynb

@@ -104,9 +104,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "scrolled": true
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "smiles = list(soldata[\"SMILES\"])\n",
@@ -234,9 +232,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "scrolled": true
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "# now get sequences\n",
@@ -270,9 +266,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "scrolled": false
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "model = tf.keras.Sequential()\n",
@@ -299,9 +293,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "scrolled": true
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "model.compile(tf.optimizers.Adam(1e-3), loss=\"mean_squared_error\")\n",
@@ -411,9 +403,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "scrolled": true
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "# Make sure SMILES doesn't contain multiple fragments\n",
@@ -434,9 +424,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "scrolled": false
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "from IPython.display import display, SVG\n",
@@ -509,9 +497,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "scrolled": false
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "# Inspect space\n",
@@ -533,9 +519,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "scrolled": false
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "fkw = {\"figsize\": (6, 4)}\n",
@@ -686,9 +670,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "scrolled": true
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "# Get subsets and calculate lime importances - subsample - get rank correlation\n",
@@ -747,9 +729,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "scrolled": false
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "# Mutation\n",
@@ -773,9 +753,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "scrolled": false
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "# Alphabet\n",

paper2_LIME/Tutorial.ipynb

@@ -145,9 +145,7 @@
    "cell_type": "code",
    "execution_count": null,
    "id": "ed252205",
-   "metadata": {
-    "scrolled": false
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "exmol.lime_explain(space)\n",

paper3_Scents/GNNModelTrainingAndEvaluation.ipynb

@@ -21,8 +21,7 @@
      "base_uri": "https://localhost:8080/"
     },
     "id": "gZeyyFPJbYxi",
-    "outputId": "2ac9b9a9-ca8b-4608-d3a4-a2c8f937f71f",
-    "scrolled": true
+    "outputId": "2ac9b9a9-ca8b-4608-d3a4-a2c8f937f71f"
    },
    "outputs": [],
    "source": [
@@ -562,9 +561,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "scrolled": false
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "# Train model\n",
@@ -1041,8 +1038,7 @@
    "cell_type": "code",
    "execution_count": null,
    "metadata": {
-    "id": "BBgB_3tPbYxx",
-    "scrolled": true
+    "id": "BBgB_3tPbYxx"
    },
    "outputs": [],
    "source": [

requirements.txt

@@ -2,5 +2,4 @@ pre-commit
 mypy
 pytest
 pytest-cov
-click
 openai

setup.py

@@ -28,7 +28,7 @@
         "skunk >= 0.4.0",
         "importlib-resources",
         "synspace",
-        "langchain",
+        "langchain==0.0.343",
     ],
     test_suite="tests",
     long_description=long_description,

tests/test_exmol.py

@@ -45,7 +45,6 @@ def test_sanitize_smiles_chiral():
     assert "@" in result[1]
 
 
-# TODO let STONED people write these when they finish their repo
 def test_run_stoned():
     result = exmol.run_stoned(
         "N#CC=CC(C(=O)NCC1=CC=CC=C1C(=O)N)(C)CC2=CC=C(F)C=C2CC",
@@ -199,14 +198,27 @@ def test_run_custom():
 
 def test_run_stones_alphabet():
     result = exmol.run_stoned(
-        "N#CC=CC(C(=O)NCC1=CC=CC=C1C(=O)N)(C)CC2=CC=C(F)C=C2CC",
-        num_samples=10,
-        max_mutations=1,
+        "C1=CC=C(C=C1)C2=CC=CC=C2",
+        num_samples=25,
+        max_mutations=3,
         alphabet=["[C]", "[O]"],
     )
     # Can get duplicates
     assert len(result[0]) >= 0
 
+    # check no other characters
+    for mol in result[0]:
+        print(["C", "O", "#", "(", ")"] + list([str(i) for i in range(10)]))
+        print(mol)
+        assert not any(
+            [
+                c
+                not in ["C", "O", "#", "(", ")", "="]
+                + list([str(i) for i in range(10)])
+                for c in mol
+            ]
+        )
+
 
 def test_sample():
     def model(s, se):