clari 0.1.0__py3-none-any.whl

clari/assets/Elemental_Radii_Alvarez.csv

@@ -0,0 +1,111 @@
+Element Name,Symbol,Period,Group,Block,Atomic Number,Atomic Weight,Covalent Radius,vdW Radius
+Actinium,Ac,7Act,20,f,89,[227],2.15,2.8
+Aluminium,Al,3,13,p,13,26.982,1.21,2.25
+Americium,Am,7Act,20,f,95,[243],1.8,2.83
+Antimony,Sb,5,15,p,51,121.76,1.39,2.47
+Argon,Ar,3,18,p,18,39.948,1.51,1.83
+Arsenic,As,4,15,p,33,74.922,1.21,1.88
+Astatine,At,6,17,p,85,[210],1.21,2
+Barium,Ba,6,2,s,56,137.327,2.15,3.03
+Berkelium,Bk,7Act,20,f,97,[247],1.54,3.4
+Beryllium,Be,2,2,s,4,9.012,0.96,1.98
+Bismuth,Bi,6,15,p,83,208.98,1.48,2.54
+Bohrium,Bh,7,7,d,107,[264],1.5,2
+Boron,B,2,13,p,5,10.811,0.83,1.91
+Bromine,Br,4,17,p,35,79.904,1.21,1.86
+Cadmium,Cd,5,12,d,48,112.411,1.54,2.49
+Caesium,Cs,6,1,s,55,132.905,2.44,3.48
+Calcium,Ca,4,2,s,20,40.078,1.76,2.62
+Californium,Cf,7Act,20,f,98,[251],1.83,3.05
+Carbon,C,2,14,p,6,12.011,0.68,1.77
+Cerium,Ce,6Lan,19,f,58,140.116,2.04,2.88
+Chlorine,Cl,3,17,p,17,35.453,0.99,1.82
+Chromium,Cr,4,6,d,24,51.996,1.39,2.45
+Cobalt,Co,4,9,d,27,58.933,1.26,2.4
+Copper,Cu,4,11,d,29,63.546,1.32,2.38
+Curium,Cm,7Act,20,f,96,[247],1.69,3.05
+Darmstadtium,Ds,7,10,d,110,[271],1.5,2
+Dubnium,Db,7,5,d,105,[262],1.5,2
+Dysprosium,Dy,6Lan,19,f,66,162.5,1.92,2.87
+Einsteinium,Es,7Act,20,f,99,[252],1.5,2.7
+Erbium,Er,6Lan,19,f,68,167.26,1.89,2.83
+Europium,Eu,6Lan,19,f,63,151.964,1.98,2.87
+Fermium,Fm,7Act,20,f,100,[257],1.5,2
+Fluorine,F,2,17,p,9,18.998,0.64,1.46
+Francium,Fr,7,1,s,87,[223],2.6,2
+Gadolinium,Gd,6Lan,19,f,64,157.25,1.96,2.83
+Gallium,Ga,4,13,p,31,69.723,1.22,2.32
+Germanium,Ge,4,14,p,32,72.61,1.17,2.29
+Gold,Au,6,11,d,79,196.967,1.36,2.32
+Hafnium,Hf,6,4,d,72,178.49,1.75,2.63
+Hassium,Hs,7,8,d,108,[269],1.5,2
+Helium,He,1,18,p,2,4.003,1.5,1.43
+Holmium,Ho,6Lan,19,f,67,164.93,1.92,2.81
+Hydrogen,H,1,1,s,1,1.008,0.23,1.2
+Indium,In,5,13,p,49,114.818,1.42,2.43
+Iodine,I,5,17,p,53,126.904,1.4,2.04
+Iridium,Ir,6,9,d,77,192.217,1.41,2.41
+Iron,Fe,4,8,d,26,55.845,1.52,2.44
+Krypton,Kr,4,18,p,36,83.8,1.5,2.25
+Lanthanum,La,6Lan,19,f,57,138.906,2.07,2.98
+Lawrencium,Lr (Lw),7,3,d,103,[262],1.5,2
+Lead,Pb,6,14,p,82,207.2,1.46,2.6
+Lithium,Li,2,1,s,3,6.941,1.28,2.12
+Lutetium,Lu,6,3,d,71,174.967,1.87,2.74
+Magnesium,Mg,3,2,s,12,24.305,1.41,2.51
+Manganese,Mn,4,7,d,25,54.938,1.61,2.45
+Meitnerium,Mt,7,9,d,109,[268],1.5,2
+Mendelevium,Md,7Act,20,f,101,[258],1.5,2
+Mercury,Hg,6,12,d,80,200.59,1.32,2.45
+Molybdenum,Mo,5,6,d,42,95.94,1.54,2.45
+Neodymium,Nd,6Lan,19,f,60,144.24,2.01,2.95
+Neon,Ne,2,18,p,10,20.18,1.5,1.58
+Neptunium,Np,7Act,20,f,93,[237],1.9,2.82
+Nickel,Ni,4,10,d,28,58.693,1.24,2.4
+Niobium,Nb,5,5,d,41,92.906,1.64,2.56
+Nitrogen,N,2,15,p,7,14.007,0.68,1.66
+Nobelium,No,7Act,20,f,102,[259],1.5,2
+Osmium,Os,6,8,d,76,190.23,1.44,2.48
+Oxygen,O,2,16,p,8,15.999,0.68,1.5
+Palladium,Pd,5,10,d,46,106.42,1.39,2.15
+Phosphorus,P,3,15,p,15,30.974,1.05,1.9
+Platinum,Pt,6,10,d,78,195.078,1.36,2.29
+Plutonium,Pu,7Act,20,f,94,[244],1.87,2.81
+Polonium,Po,6,16,p,84,[210],1.4,2
+Potassium,K,4,1,s,19,39.098,2.03,2.73
+Praseodymium,Pr,6Lan,19,f,59,140.908,2.03,2.92
+Promethium,Pm,6Lan,19,f,61,[145],1.99,2
+Protactinium,Pa,7Act,20,f,91,231.036,2,2.88
+Radium,Ra,7,2,s,88,[226],2.21,2
+Radon,Rn,6,18,p,86,[222],1.5,2
+Rhenium,Re,6,7,d,75,186.207,1.51,2.49
+Rhodium,Rh,5,9,d,45,102.906,1.42,2.44
+Rubidium,Rb,5,1,s,37,85.468,2.2,3.21
+Ruthenium,Ru,5,8,d,44,101.07,1.46,2.46
+Rutherfordium,Rf,7,4,d,104,[261],1.5,2
+Samarium,Sm,6Lan,19,f,62,150.36,1.98,2.9
+Scandium,Sc,4,3,d,21,44.956,1.7,2.58
+Seaborgium,Sg,7,6,d,106,[266],1.5,2
+Selenium,Se,4,16,p,34,78.96,1.22,1.82
+Silicon,Si,3,14,p,14,28.086,1.2,2.19
+Silver,Ag,5,11,d,47,107.868,1.45,2.53
+Sodium,Na,3,1,s,11,22.991,1.66,2.5
+Strontium,Sr,5,2,s,38,87.62,1.95,2.84
+Sulphur,S,3,16,p,16,32.066,1.02,1.89
+Tantalum,Ta,6,5,d,73,180.948,1.7,2.53
+Technetium,Tc,5,7,d,43,[98],1.47,2.44
+Tellurium,Te,5,16,p,52,127.6,1.47,1.99
+Terbium,Tb,6Lan,19,f,65,158.925,1.94,2.79
+Thallium,Tl,6,13,p,81,204.383,1.45,2.47
+Thorium,Th,7Act,20,f,90,232.038,2.06,2.93
+Thulium,Tm,6Lan,19,f,69,168.934,1.9,2.79
+Tin,Sn,5,14,p,50,118.71,1.39,2.42
+Titanium,Ti,4,4,d,22,47.867,1.6,2.46
+Tungsten,W,6,6,d,74,183.84,1.62,2.57
+Uranium,U,7Act,20,f,92,238.029,1.96,2.71
+Vanadium,V,4,5,d,23,50.942,1.53,2.42
+Xenon,Xe,5,18,p,54,131.29,1.5,2.06
+Ytterbium,Yb,6Lan,19,f,70,173.04,1.87,2.8
+Yttrium,Y,5,3,d,39,88.906,1.9,2.75
+Zinc,Zn,4,12,d,30,65.39,1.22,2.39
+Zirconium,Zr,5,4,d,40,91.224,1.75,2.52

clari/assets/posebusters_no_strain.yml

@@ -0,0 +1,112 @@
+# Setup test modules
+modules:
+  - name: "Loading"
+    function: loading
+    chosen_binary_test_output:
+      - mol_pred_loaded
+    rename_outputs:
+      mol_pred_loaded: "MOL_PRED loaded"
+
+  - name: "Chemistry"
+    function: rdkit_sanity
+    chosen_binary_test_output:
+      - passes_rdkit_sanity_checks
+    rename_outputs:
+      passes_rdkit_sanity_checks: "Sanitization"
+
+  - name: "Chemistry"
+    function: inchi_convertible
+    chosen_binary_test_output:
+      - inchi_convertible
+    rename_outputs:
+      inchi_convertible: "InChI convertible"
+
+  - name: "Chemistry"
+    function: atoms_connected
+    chosen_binary_test_output:
+      - all_atoms_connected
+    rename_outputs:
+      all_atoms_connected: "All atoms connected"
+
+  - name: "Geometry"
+    function: "distance_geometry"
+    parameters:
+      bound_matrix_params:
+        set15bounds: True # topology based bounds also for 1,5- not just until 1,4-
+        scaleVDW: True # scale down lower bounds for atoms less than 5 bonds apart
+        doTriangleSmoothing: True
+        useMacrocycle14config: False
+      threshold_bad_bond_length: 0.25 # widens DG bound by this factor
+      threshold_bad_angle: 0.25 # widens DG bound by this factor
+      threshold_clash: 0.3 # widens DG bound by this factor
+      ignore_hydrogens: False # ignore hydrogens
+      sanitize: True # sanitize molecule before running DG module (recommended)
+    chosen_binary_test_output:
+      - bond_lengths_within_bounds
+      - bond_angles_within_bounds
+      - no_internal_clash
+    rename_outputs:
+      bond_lengths_within_bounds: "Bond lengths"
+      bond_angles_within_bounds: "Bond angles"
+      no_internal_clash: "Internal steric clash"
+
+  - name: "Ring flatness"
+    function: "flatness"
+    parameters:
+      flat_systems: # list atoms which together should lie on plane as SMARTS matches
+        aromatic_5_membered_rings_sp2: "[ar5^2]1[ar5^2][ar5^2][ar5^2][ar5^2]1"
+        aromatic_6_membered_rings_sp2: "[ar6^2]1[ar6^2][ar6^2][ar6^2][ar6^2][ar6^2]1"
+      threshold_flatness: 0.25 # max distance in A to closest shared plane
+    chosen_binary_test_output:
+      - flatness_passes
+    rename_outputs:
+      num_systems_checked: number_aromatic_rings_checked
+      num_systems_passed: number_aromatic_rings_pass
+      max_distance: aromatic_ring_maximum_distance_from_plane
+      flatness_passes: "Aromatic ring flatness"
+
+  - name: "Double bond flatness"
+    function: "flatness"
+    parameters:
+      flat_systems: # list atoms which together should lie on plane as SMARTS matches
+        trigonal_planar_double_bonds: "[C;X3;^2](*)(*)=[C;X3;^2](*)(*)"
+      threshold_flatness: 0.25 # max distance in A to closest shared plane
+    chosen_binary_test_output:
+      - flatness_passes
+    rename_outputs:
+      num_systems_checked: number_double_bonds_checked
+      num_systems_passed: number_double_bonds_pass
+      max_distance: double_bond_maximum_distance_from_plane
+      flatness_passes: "Double bond flatness"
+
+  # - name: "Energy ratio"
+  #   function: energy_ratio
+  #   parameters:
+  #     threshold_energy_ratio: 100.0
+  #     ensemble_number_conformations: 50
+  #     inchi_strict: False
+  #   chosen_binary_test_output:
+  #     - energy_ratio_passes
+  #   rename_outputs:
+  #     energy_ratio_passes: "Internal energy"
+
+# Options for loading molecule files with RDKit
+loading:
+  mol_pred:
+    cleanup: False
+    sanitize: False
+    add_hs: False
+    assign_stereo: False
+    load_all: True
+  mol_true:
+    cleanup: False
+    sanitize: False
+    add_hs: False
+    assign_stereo: False
+    load_all: True
+  mol_cond:
+    cleanup: False
+    sanitize: False
+    add_hs: False
+    assign_stereo: False
+    proximityBonding: False

clari/chem/__init__.py

@@ -0,0 +1,10 @@
+from clari.chem.common import (
+    BOND_TO_INDEX,
+    INDEX_TO_BOND,
+    PTABLE,
+    distance_lbound,
+    element_radii,
+    silenced_rdlogger,
+)
+from clari.chem.crystal import Crystal
+from clari.chem.featurize import ATOM_FEATURES, featurize

clari/chem/common.py

@@ -0,0 +1,79 @@
+import contextlib
+import pathlib
+
+import polars as pl
+import torch
+from rdkit import Chem, RDLogger
+from rdkit.Chem.rdchem import GetPeriodicTable
+
+from clari.paths import ASSETS_DIR
+
+PTABLE = GetPeriodicTable()
+
+
+# 0 means no bonds!
+INDEX_TO_BOND = {
+    1: Chem.BondType.SINGLE,
+    2: Chem.BondType.DOUBLE,
+    3: Chem.BondType.TRIPLE,
+    4: Chem.BondType.AROMATIC,
+    5: Chem.BondType.UNSPECIFIED,
+}
+
+BOND_TO_INDEX = [5] * 22
+for i, btype in INDEX_TO_BOND.items():
+    BOND_TO_INDEX[int(btype)] = i
+
+
+@contextlib.contextmanager
+def silenced_rdlogger():
+    logger = RDLogger.logger()
+    logger.setLevel(RDLogger.CRITICAL)
+    try:
+        yield
+    finally:
+        logger.setLevel(RDLogger.INFO)
+
+
+def xyzfile(atoms, coords):
+    file = [f"{len(atoms)}\n"]
+    for a, p in zip(atoms, coords, strict=False):
+        symbol = PTABLE.GetElementSymbol(int(a))
+        x, y, z = p.tolist()
+        file.append(f"{symbol} {x:f} {y:f} {z:f}")
+    return "\n".join(file)
+
+
+# Downloaded from: https://www.ccdc.cam.ac.uk/media/Elemental_Radii_Alvarez.xlsx
+def load_radii():
+    radii = {"vdw": torch.full([118], 2.0), "cov": torch.full([118], 1.5)}
+    df = pl.read_csv(pathlib.Path(ASSETS_DIR / "Elemental_Radii_Alvarez.csv"))
+    for row in df.iter_rows(named=True):
+        radii["vdw"][row["Atomic Number"]] = row["vdW Radius"]
+        radii["cov"][row["Atomic Number"]] = row["Covalent Radius"]
+    return radii
+
+
+RADII_CACHE = load_radii()
+
+
+def element_radii(z, rtype):
+    radii = RADII_CACHE[rtype]
+    if isinstance(z, int):
+        r = radii[z].item()
+    else:
+        r = radii.to(z)[z]
+    return r
+
+
+def distance_lbound(z1, z2, bond_mask):
+    nonmetals = torch.tensor([1, 2, 6, 7, 8, 9, 10, 15, 16, 17, 18, 34, 35, 36, 53, 54]).to(z1)
+    is_metal1 = ~torch.isin(z1, nonmetals)
+    is_metal2 = ~torch.isin(z2, nonmetals)
+    metal_mask = is_metal1.unsqueeze(-1) | is_metal2.unsqueeze(-2)
+
+    r1 = element_radii(z1, "cov")
+    r2 = element_radii(z2, "cov")
+    lb = r1.unsqueeze(-1) + r2.unsqueeze(-2)
+    lb = torch.clip(torch.where(bond_mask | metal_mask, 0.6, 1.0) * lb, min=0.5)
+    return lb