Source code for swiftpol.build

#Init
import rdkit
from rdkit import Chem
from rdkit.Chem import AllChem
import os
import random
import numpy as np
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import time
from collections import Counter
try:
    from openeye import oechem
    oechem_imported = True
except:
    import warnings
    warnings.warn("OpenEye is not installed. You will not be able to use OpenEye Toolkits for conformer generation.")
    oechem_imported = False

import openmm
from openff.toolkit.topology import Molecule, Topology
from openff.toolkit.typing.engines.smirnoff import ForceField
from openff.toolkit.utils import get_data_file_path
from openff.toolkit.utils.toolkits import RDKitToolkitWrapper, AmberToolsToolkitWrapper
from openff.units import unit
from pandas import read_csv


from openff.interchange import Interchange
from openff.interchange.components._packmol import UNIT_CUBE, pack_box, _max_dist_between_points


from functools import reduce
from statistics import mean
from rdkit.Chem.Descriptors import ExactMolWt
from openff.interchange import Interchange


#Build polymer - generic
[docs] def build_polymer(sequence, monomer_list, reaction, terminal='hydrogen', chain_num=1): """ Constructs a polymer from a given sequence of monomers. Parameters ---------- sequence : str A string representing the sequence of monomers (e.g., 'ABAB'). monomer_list : list A list of SMILES strings representing the monomers. reaction : rdkit.Chem.rdChemReactions.ChemicalReaction An RDKit reaction object used to link monomers. terminal : str, optional The terminal group to be added to the polymer. Options are 'hydrogen', 'carboxyl', 'ester', or a canonical smiles string to insert as the terminal. Default is 'hydrogen'. chain_number : int, optional The number of polymer chains to construct. Default is 1. Input used for ensemble build. Returns ------- rdkit.Chem.rdchem.Mol The constructed polymer as an RDKit molecule object. """ from rdkit import RDLogger RDLogger.DisableLog('rdApp.*') monomers = {} for x in sorted(list(set(sequence))): ind = sorted(list(set(sequence))).index(x) monomers[x] = monomer_list[ind] hits = Chem.MolFromSmiles(monomers[sequence[0]]).GetSubstructMatches(Chem.MolFromSmarts('I')) mw = Chem.RWMol(Chem.MolFromSmiles(monomers[sequence[0]])) mw.ReplaceAtom(hits[0][0],Chem.Atom(17)) Chem.SanitizeMol(mw) mw.CommitBatchEdit() polymer = Chem.AddHs(mw) info = Chem.AtomPDBResidueInfo() info.SetResidueName(str(chain_num) + sequence[0] + str(1)) info.SetResidueNumber(1) [atom.SetMonomerInfo(info) for atom in polymer.GetAtoms()] Chem.SanitizeMol(polymer) for i in range(len(sequence))[1:]: if sequence[i] == 'A': A = Chem.MolFromSmiles(monomers['A']) A = Chem.AddHs(A) info = Chem.AtomPDBResidueInfo() info.SetResidueName(str(chain_num) + 'A' + str(i+1)) info.SetResidueNumber(i+1) [atom.SetMonomerInfo(info) for atom in A.GetAtoms()] try: polymer = reaction.RunReactants((polymer, A))[0][0] except IndexError: raise ValueError("Reaction failed. Please check the reaction SMARTS and monomer SMILES. For support with constructing reaction SMARTS, raise an issue at https://github.com/matta-research-group/SwiftPol/issues") Chem.SanitizeMol(polymer) elif sequence[i] == 'B': B = Chem.MolFromSmiles(monomers['B']) B = Chem.AddHs(B) info = Chem.AtomPDBResidueInfo() info.SetResidueName(str(chain_num) + 'B' + str(i+1)) info.SetResidueNumber(i+1) [atom.SetMonomerInfo(info) for atom in B.GetAtoms()] try: polymer = reaction.RunReactants((polymer, B))[0][0] except IndexError: raise ValueError("Reaction failed. Please check the reaction SMARTS and monomer SMILES. For support with constructing reaction SMARTS, raise an issue at https://github.com/matta-research-group/SwiftPol/issues") Chem.SanitizeMol(polymer) elif sequence[i] == 'S': S = Chem.MolFromSmiles(monomers['S']) S = Chem.AddHs(S) info = Chem.AtomPDBResidueInfo() info.SetResidueName(str(chain_num) + 'S' + str(i+1)) info.SetResidueNumber(i+1) [atom.SetMonomerInfo(info) for atom in S.GetAtoms()] try: polymer = reaction.RunReactants((polymer, S))[0][0] except IndexError: raise ValueError("Reaction failed. Please check the reaction SMARTS and monomer SMILES. For support with constructing reaction SMARTS, raise an issue at https://github.com/matta-research-group/SwiftPol/issues") Chem.SanitizeMol(polymer) if terminal == 'hydrogen': hydrogen = Chem.MolFromSmiles('[H]') info = Chem.AtomPDBResidueInfo() info.SetResidueName(str(chain_num) + sequence[0] + str(1)) info.SetResidueNumber(1) [atom.SetMonomerInfo(info) for atom in hydrogen.GetAtoms()] polymer = Chem.ReplaceSubstructs(polymer, Chem.MolFromSmarts('Cl'), hydrogen)[0] Chem.AddHs(polymer) elif terminal == 'carboxyl': carboxyl = Chem.MolFromSmiles('C(=O)[OH]') info = Chem.AtomPDBResidueInfo() info.SetResidueName(str(chain_num) + sequence[0] + str(1)) info.SetResidueNumber(1) [atom.SetMonomerInfo(info) for atom in carboxyl.GetAtoms()] polymer = Chem.ReplaceSubstructs(polymer, Chem.MolFromSmarts('Cl'), carboxyl)[0] elif terminal == 'ester': carbon = Chem.MolFromSmiles('[CH3]') carbon = Chem.AddHs(carbon) info = Chem.AtomPDBResidueInfo() info.SetResidueName(str(chain_num) + sequence[0] + str(1)) info.SetResidueNumber(1) [atom.SetMonomerInfo(info) for atom in carbon.GetAtoms()] polymer = Chem.ReplaceSubstructs(polymer, Chem.MolFromSmarts('Cl'), carbon)[0] #Chem.AddHs(polymer) polymer = Chem.ReplaceSubstructs(polymer, Chem.MolFromSmarts('Cl'), Chem.MolFromSmiles('C'))[0] else: try: term = Chem.MolFromSmiles(terminal) except: raise ValueError("Terminal must be a valid SMILES string or one of the following options: 'hydrogen', 'carboxyl', 'ester'.") info = Chem.AtomPDBResidueInfo() info.SetResidueName(str(chain_num) + sequence[0] + str(1)) info.SetResidueNumber(1) [atom.SetMonomerInfo(info) for atom in term.GetAtoms()] polymer = Chem.ReplaceSubstructs(polymer, Chem.MolFromSmarts('Cl'), term)[0] hydrogen = Chem.MolFromSmiles('[H]') info = Chem.AtomPDBResidueInfo() info.SetResidueName(str(chain_num) + sequence[-1] + str(len(sequence))) info.SetResidueNumber(len(sequence)) [atom.SetMonomerInfo(info) for atom in hydrogen.GetAtoms()] polymer = Chem.ReplaceSubstructs(polymer, Chem.MolFromSmarts('I'), hydrogen)[0] #remove any excess iodine Chem.SanitizeMol(polymer) return polymer
[docs] def build_linear_copolymer(sequence, monomer_a_smiles, monomer_b_smiles, reaction=AllChem.ReactionFromSmarts('[C:1][HO:2].[HO:3][C:4]>>[C:1][O:2][C:4].[O:3]')): """ Constructs a linear co-polymer from the provided sequence of monomers. This function takes a sequence of monomers represented as 'A' and 'B', and the SMILES strings of two monomers. It constructs a co-polymer based on the sequence, using the provided reaction SMARTS for joining the monomers. The function returns the sanitized polymer and the percentage composition of each monomer in the polymer. Parameters ---------- sequence : str A string representing the sequence of monomers. 'A' represents monomer_a and 'B' represents monomer_b. monomer_a_smiles : str The SMILES string of monomer A. monomer_b_smiles : str The SMILES string of monomer B. reaction : rdkit.Chem.rdChemReactions.ChemicalReaction, optional The reaction SMARTS used for joining the monomers. Defaults to '[C:1][HO:2].[HO:3][C:4]>>[C:1][O:2][C:4].[O:3]', representing a condensation polymerisation. Returns ------- tuple A tuple containing the following elements: - sanitized_polymer (rdkit.Chem.rdchem.Mol): The constructed and sanitized polymer. - percentage_monomer_a (float): The percentage composition of monomer A in the polymer. - percentage_monomer_b (float): The percentage composition of monomer B in the polymer. """ # Initialize the polymer with an iodine blocker polymer = Chem.MolFromSmiles('OC(=O)I') A_count=0 B_count=0 A = Chem.MolFromSmiles(monomer_a_smiles) B = Chem.MolFromSmiles(monomer_b_smiles) # Build the polymer based on the sequence for i in range(len(sequence)): if sequence[i] == 'A': polymer = reaction.RunReactants((polymer, A))[0][0] Chem.SanitizeMol(polymer) A_count+=1 elif sequence[i] == 'B': polymer = reaction.RunReactants((polymer, B))[0][0] Chem.SanitizeMol(polymer) B_count+=1 # Calculate the percentage composition of each monomer A_ratio = round((A_count/len(sequence))*100,2) B_ratio = round((B_count/len(sequence))*100,2) # Remove the iodine blocker polymer = Chem.ReplaceSubstructs(polymer, Chem.MolFromSmarts('OC(=O)I'), Chem.MolFromSmarts('[O]'))[0] Chem.SanitizeMol(polymer) return polymer, A_ratio, B_ratio
[docs] def PDI(chains): """ Calculates the Polydispersity Index (PDI), number-average molecular weight (Mn), and weight-average molecular weight (Mw) of a list of chains. Parameters ---------- chains : list A list of molecular chains. Each chain is represented as an RDKit molecule object. Returns ------- tuple A tuple containing the following elements: - PDI (float): The Polydispersity Index, which is the ratio of Mw to Mn. - Mn (float): The number-average molecular weight. - Mw (float): The weight-average molecular weight. """ # Calculate the molecular weights of the chains mw_list = [ExactMolWt(chain) for chain in chains] # Count instances of each molecular weight mw_counts = Counter(mw_list) # Calculate Mn total_chains = sum(mw_counts.values()) # Total number of chains Mn = sum(mw * count for mw, count in mw_counts.items()) / total_chains # Calculate Mw Mw = sum(mw**2 * count for mw, count in mw_counts.items()) / sum(mw * count for mw, count in mw_counts.items()) # Calculate PDI - Polydispersity Index PDI = Mw / Mn return PDI, Mn, Mw
[docs] def blockiness_gen(sequence, wrt='A'): """ Calculate the blockiness and average block length of a co-polymer sequence. This function calculates the blockiness of a co-polymer sequence by counting the occurrences of 'BB' and 'BA' or 'AB' in the sequence. It also calculates the average block length of 'A' and 'B' monomers in the sequence. Parameters ---------- sequence : str A string representing the co-polymer sequence. 'A' represents one type of monomer and 'B' represents another type. Returns ------- tuple A tuple containing the following elements: - blockiness (float): The blockiness of the co-polymer sequence. Calculated as the ratio of 'BB' to 'BA' or 'AB'. - block_length_A (float): The average block length of 'A' in the sequence. - block_length_B (float): The average block length of 'B' in the sequence. Notes ----- If the sequence does not contain both 'A' and 'B', the function returns a string indicating that the molecule is not a co-polymer. """ if wrt == 'A': if 'A' in sequence and 'B' in sequence: #Check if sequence is a co-polymer AB = sequence.count('AB') BB = sequence.count('BB') BA = sequence.count('BA') AA = sequence.count('AA') if 'BA' in sequence: blockiness = AA/BA else: blockiness = AA/AB #Calculate block length B block_list_B = [x for x in sequence.split('A') if x!=''] block_length_B = mean([len(b) for b in block_list_B]) #Calculate block length A block_list_A = [x for x in sequence.split('B') if x!=''] block_length_A = mean([len(b) for b in block_list_A]) return blockiness, block_length_B, block_length_A else: return 'Molecule is not a co-polymer, no blockiness calculation performed', 0, len(sequence) elif wrt == 'B': if 'A' in sequence and 'B' in sequence: AB = sequence.count('AB') BB = sequence.count('BB') BA = sequence.count('BA') AA = sequence.count('AA') if 'AB' in sequence: blockiness = BB/AB else: blockiness = BB/BA #Calculate block length A block_list_A = [x for x in sequence.split('B') if x!=''] block_length_A = mean([len(b) for b in block_list_A]) #Calculate block length B block_list_B = [x for x in sequence.split('A') if x!=''] block_length_B = mean([len(b) for b in block_list_B]) return blockiness, block_length_A, block_length_B else: return 'Molecule is not a co-polymer, no blockiness calculation performed', 0, len(sequence) else: raise ValueError("wrt parameter must be 'A' or 'B'")
[docs] def calculate_box_components(chains, monomers, sequence, salt_concentration = 0.0* unit.mole / unit.liter, residual_monomer = 0.00): """ Calculates the components required to construct a simulation box for a given set of molecular chains. This function determines the quantity of each molecule type required, considering the salt concentration and residual monomer concentration. It is adapted from the OpenFF Toolkit Packmol wrapper's solvate_topology function. Parameters: ----------- chains : list A list of molecular chains to be included in the simulation box. sequence : str A string representing the sequence of the molecular chains. 'G' and 'L' represent different types of monomers. salt_concentration : float, optional The desired salt concentration in the simulation box. Defaults to 0 M. residual_monomer : float, optional The desired residual monomer concentration in the simulation box. Defaults to 0.00%. Returns: -------- tuple A tuple containing the following elements: - molecules (list): A list of molecules to be included in the simulation box. - number_of_copies (list): A list indicating the quantity of each molecule to be included in the simulation box. - topology (openff.toolkit.topology.Topology): The topology of the simulation box. - box_vectors (numpy.ndarray): The vectors defining the dimensions of the simulation box. Notes: ------ This function is adapted from the OpenFF Toolkit Packmol wrapper's solvate_topology function. """ from openff.toolkit.topology import Molecule, Topology from openff.interchange.components._packmol import UNIT_CUBE, pack_box, RHOMBIC_DODECAHEDRON, solvate_topology from openff.interchange.components._packmol import _max_dist_between_points, _compute_brick_from_box_vectors, _center_topology_at import warnings #Create molecules for the purpose of mass calculation #Water water = Molecule.from_smiles('O') water.generate_unique_atom_names() water.generate_conformers() water_mass = sum([atom.mass for atom in water.atoms]) #Sodium na = Molecule.from_smiles('[Na+]') na.generate_unique_atom_names() na.generate_conformers() #Chloride cl = Molecule.from_smiles('[Cl-]') cl.generate_unique_atom_names() cl.generate_conformers() nacl_mass = sum([atom.mass for atom in na.atoms]) + sum([atom.mass for atom in cl.atoms]) # Create a topology from the chains topology = Topology.from_molecules(chains) nacl_conc=salt_concentration padding= 1.0 * unit.nanometer box_shape= UNIT_CUBE target_density= 1.0 * unit.gram / unit.milliliter # Compute box vectors from the solute length and requested padding if chains[0].n_conformers == 0: raise ValueError("The solvate_topology function requires that the solute has at least one conformer.") solute_length = max(_max_dist_between_points(chains[i].to_topology().get_positions()) for i in range(len(chains))) image_distance = solute_length + padding * 2 box_vectors = box_shape * image_distance # Compute target masses of solvent box_volume = np.linalg.det(box_vectors.m) * box_vectors.u**3 target_mass = box_volume * target_density solvent_mass = target_mass - sum(sum([atom.mass for atom in molecule.atoms]) for molecule in topology.molecules) # Compute the number of NaCl to add from the mass and concentration nacl_mass_fraction = (nacl_conc * nacl_mass) / (55.5 * unit.mole / unit.liter * water_mass) nacl_to_add = ((solvent_mass * nacl_mass_fraction) / nacl_mass).m_as(unit.dimensionless).round() water_to_add = int(round((solvent_mass) / water_mass).m_as(unit.dimensionless).round()) # Neutralise the system by adding and removing salt solute_charge = sum([molecule.total_charge for molecule in topology.molecules]) na_to_add = int(round(np.ceil(nacl_to_add - solute_charge.m / 2.0))) cl_to_add = int(round(np.floor(nacl_to_add + solute_charge.m / 2.0))) rolling_mass=0 for m in topology.molecules: rolling_mass += sum(atom.mass for atom in m.atoms) rolling_mass += nacl_mass * nacl_to_add # Calculate residual monomer to add mass_to_add = (rolling_mass.magnitude/100-residual_monomer) * residual_monomer if mass_to_add < 90: warnings.warn('Residual monomer calculation may not be accurate for small systems, as the residual monomer mass can be lower than the weight of a single monomer. Please check the output by querying the value of residual_monomer_actual') if 'A' in sequence and 'B' in sequence: A_rd = Chem.MolFromSmiles(monomers[0]) A_rd = Chem.AddHs(A_rd) info = Chem.AtomPDBResidueInfo() info.SetResidueName('A' + str(1)) info.SetResidueNumber(1) [atom.SetMonomerInfo(info) for atom in A_rd.GetAtoms()] A = Molecule.from_rdkit(A_rd) A.name = 'A_residual' A_mass = sum([atom.mass for atom in A.atoms]) B_rd = Chem.MolFromSmiles(monomers[1]) B_rd = Chem.AddHs(B_rd) info = Chem.AtomPDBResidueInfo() info.SetResidueName('B' + str(1)) info.SetResidueNumber(1) [atom.SetMonomerInfo(info) for atom in B_rd.GetAtoms()] B = Molecule.from_rdkit(B_rd) B_mass = sum([atom.mass for atom in B.atoms]) B.name = 'B_residual' for r in range(0,100): if (r * A_mass.magnitude) + (r * B_mass.magnitude) <= mass_to_add: A_to_add = r B_to_add = r else: break residual_monomer_actual = ((A_to_add * A_mass.magnitude + B_to_add * B_mass.magnitude) / rolling_mass.magnitude) *100 molecules = [water, na, cl, A, B] number_of_copies=[water_to_add, na_to_add, cl_to_add, A_to_add, B_to_add] elif 'A' in sequence and 'B' not in sequence: A_rd = Chem.MolFromSmiles(monomers[0]) A_rd = Chem.AddHs(A_rd) info = Chem.AtomPDBResidueInfo() info.SetResidueName('A' + str(1)) info.SetResidueNumber(1) [atom.SetMonomerInfo(info) for atom in A_rd.GetAtoms()] A = Molecule.from_rdkit(A_rd) A_mass = sum([atom.mass for atom in A.atoms]) A.name = 'A_residual' B = Molecule.from_smiles('C') B.name = 'B_residual' for r in range(0,100): if r * A_mass.magnitude <= mass_to_add: A_to_add = r else: break B_to_add = 0 residual_monomer_actual = ((A_to_add * A_mass.magnitude) / rolling_mass.magnitude) * 100 molecules = [water, na, cl, A, B] number_of_copies=[water_to_add, na_to_add, cl_to_add, A_to_add, B_to_add] return molecules, number_of_copies, topology, box_vectors, residual_monomer_actual
[docs] def introduce_stereoisomers(stereo_monomer, instance, seq): """ Introduce stereoisomers into a polymer sequence by replacing a specified percentage of 'A' monomers with 'S'. This function replaces a specified percentage of 'A' monomers with 'S' in the given sequence. The replacements are made in pairs of 'A' monomers to ensure stereoisomerism. Parameters ---------- stereo_monomer : str The monomer to be replaced with its stereoisomer (e.g., 'A'). instance : float The fraction of 'stereo_monomer' monomers to be replaced with stereoisomer (e.g., 0.5 for 50%). seq : str The original polymer sequence. Returns ------- str The modified sequence with stereoisomers introduced. """ assert stereo_monomer in seq, f"Monomer {stereo_monomer} not found in sequence" assert instance <= 1 and instance > 0, "Instance must be a fraction between 0 and 1" assert instance > 0, "Instance must be greater than 0" assert type(stereo_monomer) == str, "Monomer must be a string" num_replacements = round(seq.count(stereo_monomer) * instance) seq_list = list(seq) # Initialize a counter for replacements replacement_count = 0 i = 0 while i < len(seq_list) - 1: if seq_list[i] == 'A' and seq_list[i + 1] == 'A': seq_list[i + 1] = 'S' replacement_count += 1 if replacement_count >= num_replacements: break i += 1 i += 1 modified_seq = ''.join(seq_list) return modified_seq
[docs] class polymer_system: try: from openeye import oechem except: import warnings warnings.warn("OpenEye is not installed. You will not be able to use OpenEye Toolkits for conformer generation.") from openff.toolkit.utils.toolkits import RDKitToolkitWrapper, OpenEyeToolkitWrapper from functools import reduce from statistics import mean from rdkit.Chem.Descriptors import ExactMolWt from openff.interchange import Interchange from openff.interchange.components._packmol import UNIT_CUBE, pack_box from swiftpol.build import build_polymer, PDI, blockiness_gen, calculate_box_components, introduce_stereoisomers from openff.units import unit from rdkit.Chem import AllChem
[docs] def __init__(self, monomer_list, reaction, length_target, num_chains, stereoisomerism_input=None, terminals='hydrogen', perc_A_target=100, blockiness_target=[1.0, 'A'], copolymer=False, acceptance = 10 ): """ Initialize the polymer system and build the polymer chains. **Parameters:** ------------ monomer_list (list): List of monomers to be used in the polymerization. reaction (str): The type of reaction to be used for polymerization. length_target (float): The target length of the polymer chains. num_chains (int): The number of polymer chains to be generated. stereoisomerism_input (tuple, optional): A tuple containing the monomer, instance fraction (e.g. 0.5 for 50% stereoisomer), and SMILES string of the stereoisomer to be introduced. Default is None. terminals (str, optional): The type of terminal groups to be used. Default is 'hydrogen', adds a hydrogen atom. perc_A_target (float, optional): The target percentage of monomer A in the copolymer. Default is 100. blockiness_target (float, optional): The target blockiness of the copolymer, and indication of method calculation. Default is 1.0, with reference to 'A' monomer linkages. copolymer (bool, optional): Flag to indicate if the system is a copolymer. Default is False. acceptance = % deviation of blockiness and A percentage from target values. Default is 10% **Attributes:** --------------- length_target (float): The target length of the polymer chains. terminals (str): The type of terminal groups used. blockiness_target (float): The target blockiness of the copolymer. A_target (float): The target percentage of monomer A in the copolymer. chains (list): List of polymer chains as OpenFF Molecule objects. chain_rdkit (list): List of polymer chains as RDKit molecule objects. lengths (list): List of lengths of the polymer chains. perc_A_actual (list): List of actual percentages of monomer A in the polymer chains. B_block_length (float): The average block length of monomer B in the copolymer. A_block_length (float): The average block length of monomer A in the copolymer. blockiness_list (list): List of blockiness values for the polymer chains. mean_blockiness (float): The mean blockiness of the polymer chains. mol_weight_average (float): The average molecular weight of the polymer chains. PDI (float): The polydispersity index of the polymer chains. Mn (float): The number-average molecular weight of the polymer chains. Mw (float): The weight-average molecular weight of the polymer chains. num_chains (int): The number of polymer chains generated. length_average (float): The average length of the polymer chains. min_length (float): The minimum length of the polymer chains. max_length (float): The maximum length of the polymer chains. """ self.length_target = length_target self.terminals = terminals perc_A_actual = [] if copolymer==True: self.blockiness_target = blockiness_target[0] self.A_target = perc_A_target def spec(sequence): #Define limits of A percentage and blockiness from input acceptance_dec = acceptance/100 actual_A = (sequence.count('A')/len(sequence))*100 blockiness = blockiness_gen(sequence, blockiness_target[1])[0] return actual_A > perc_A_target*(1-acceptance_dec) and actual_A < perc_A_target*(1+acceptance_dec) and blockiness>blockiness_target[0]*(1-acceptance_dec) and blockiness<blockiness_target[0]*(1+acceptance_dec) blockiness_list = [] out_of_spec = 0 BBL = [] ABL = [] chains = [] chains_rdkit = [] lengths = [] self.monomers = [mono.replace("[I]", "") for mono in monomer_list] self.reaction = reaction reaction = AllChem.ReactionFromSmarts(reaction) if stereoisomerism_input is not None: stereo_monomer, instance, new_smiles = stereoisomerism_input monomer_list.append(new_smiles) #First round of building - copolymer if copolymer==True: for n in range(num_chains): sigma = np.sqrt(np.log(1.5*(1+acceptance/100))) mu = np.log(length_target) - 0.5 * sigma**2 length_actual = np.random.lognormal(mu, sigma) sequence = reduce(lambda x, y: x + y, np.random.choice(['A', 'B'], size=(int(length_actual),), p=[perc_A_target/100,1-(perc_A_target/100)])) blockiness = blockiness_gen(sequence, blockiness_target[1])[0] if spec(sequence)==True: if stereoisomerism_input is not None: sequence_stereo = introduce_stereoisomers(stereo_monomer, instance, sequence) pol = build_polymer(sequence=sequence_stereo, monomer_list = monomer_list, reaction = reaction, terminal=terminals, chain_num=n+1) else: pol = build_polymer(sequence=sequence, monomer_list = monomer_list, reaction = reaction, terminal=terminals, chain_num=n+1) lengths.append(int(length_actual)) chains_rdkit.append(pol) chain = Molecule.from_rdkit(pol) chains.append(chain) perc_A_actual.append((sequence.count('A')/len(sequence))*100) blockiness_list.append(blockiness) BBL.append(blockiness_gen(sequence, blockiness_target[1])[1]) ABL.append(blockiness_gen(sequence, blockiness_target[1])[2]) else: out_of_spec +=1 #Second round of building while out_of_spec >0: sigma = np.sqrt(np.log(1.5*(1+acceptance/100))) mu = np.log(length_target) - 0.5 * sigma**2 length_actual = np.random.lognormal(mu, sigma) sequence = reduce(lambda x, y: x + y, np.random.choice(['A', 'B'], size=(int(length_actual),), p=[perc_A_target/100,1-(perc_A_target/100)])) blockiness = blockiness_gen(sequence, blockiness_target[1])[0] if spec(sequence)==True: if stereoisomerism_input is not None: sequence_stereo = introduce_stereoisomers(stereo_monomer, instance, sequence) pol = build_polymer(sequence=sequence_stereo, monomer_list = monomer_list, reaction = reaction, terminal=terminals, chain_num=n+1) else: pol = build_polymer(sequence=sequence, monomer_list = monomer_list, reaction = reaction, terminal=terminals, chain_num=n+1) lengths.append(int(length_actual)) chains_rdkit.append(pol) chain = Molecule.from_rdkit(pol) chains.append(chain) perc_A_actual.append((sequence.count('A')/len(sequence))*100) blockiness_list.append(blockiness) BBL.append(blockiness_gen(sequence, blockiness_target[1])[1]) ABL.append(blockiness_gen(sequence, blockiness_target[1])[2]) out_of_spec-=1 self.B_block_length = mean(BBL) self.A_block_length = mean(ABL) self.blockiness_list = blockiness_list self.mean_blockiness = mean(blockiness_list) self.perc_A_actual = perc_A_actual self.A_actual = mean(perc_A_actual) else: for n in range(num_chains): sigma = np.sqrt(np.log(1.5*(1+acceptance/100))) mu = np.log(length_target) - 0.5 * sigma**2 length_actual = np.random.lognormal(mu, sigma) sequence = reduce(lambda x, y: x + y, np.random.choice(['A', 'B'], size=(int(length_actual),), p=[perc_A_target/100,1-(perc_A_target/100)])) if stereoisomerism_input is not None: sequence_stereo = introduce_stereoisomers(stereo_monomer, instance, sequence) pol = build_polymer(sequence=sequence_stereo, monomer_list = monomer_list, reaction = reaction, terminal=terminals, chain_num=n+1) else: pol = build_polymer(sequence=sequence, monomer_list = monomer_list, reaction = reaction, terminal=terminals, chain_num=n+1) lengths.append(int(length_actual)) chains_rdkit.append(pol) chain = Molecule.from_rdkit(pol) chains.append(chain) perc_A_actual.append((sequence.count('A')/len(sequence))*100) self.A_target = perc_A_target self.B_block_length = None self.A_block_length = None self.blockiness_list = None self.mean_blockiness = None self.perc_A_actual = None self.A_actual = None self.sequence = sequence self.chains = chains for i in range(len(self.chains)): self.chains[i].name = 'chain' + str(i+1) self.chain_rdkit = chains_rdkit self.mol_weight_average = round(mean([ExactMolWt(c) for c in chains_rdkit]),2) self.PDI, self.Mn, self.Mw = PDI(chains_rdkit) self.num_chains = len(chains) self.A_actual = mean(perc_A_actual) self.perc_A_actual = perc_A_actual self.length_average = mean(lengths) self.lengths = lengths self.min_length = min(lengths) self.max_length = max(lengths) print('System built!, size =', self.num_chains)
def __repr__(self): description = (f"SwiftPol ensemble of size {self.num_chains}, " f"average chain length = {self.length_average}-mers, PDI = {self.PDI}") return description
[docs] def generate_conformers(self): """ Generate conformers for each polymer chain in the system. This method uses the OpenFF toolkit OpenEye Wrapper to generate conformers for each polymer chain in the system. It first checks if the OpenEye toolkit is licensed and available. If it is, it uses the OpenEyeToolkitWrapper to generate conformers. Otherwise, it falls back to using the RDKitToolkitWrapper. Each chain is processed to generate a single conformer, and unique atom names are assigned to each chain. Raises ------ ImportError If neither RDKit nor OpenEye toolkits are available. """ from openff.toolkit.utils.toolkits import RDKitToolkitWrapper, OpenEyeToolkitWrapper # Generate conformers using OpenFF toolkit wrapper for chain in self.chains: num = self.chains.index(chain) object = RDKitToolkitWrapper() if oechem_imported: if oechem.OEChemIsLicensed(): object = OpenEyeToolkitWrapper() object.generate_conformers(molecule=chain, n_conformers=1) chain.generate_unique_atom_names() self.chains[num] = chain
[docs] def charge_system(self, charge_scheme): """ Assign partial charges to each polymer chain in the system. This method uses one of AM1-BCC, Espaloma, or OpenFF NAGL to assign partial charges to each polymer chain in the system. It iterates over each chain in the `self.chains` list and assigns partial charges to the chain. Parameters ---------- charge_scheme : str The charge assignment scheme to use. Options are 'AM1_BCC', 'espaloma', or 'NAGL'. Raises ------ ImportError If the selected toolkit is not available. """ from swiftpol.parameterize import charge_openff_polymer for chain in self.chains: chain.partial_charges = charge_openff_polymer(chain, charge_scheme)
[docs] def export_to_csv(self, filename): """ Export all the instances in polymer_system.__init__() into a pandas DataFrame and save it as a CSV file. Parameters ---------- filename : str The name of the CSV file to save the data. """ data = { 'monomers': [self.monomers], 'length_target': [self.length_target], 'terminals': [self.terminals], 'num_chains': [self.num_chains], 'mol_weight_average': [self.mol_weight_average], 'PDI': [self.PDI], 'Mn': [self.Mn], 'Mw': [self.Mw], 'sequence': [self.sequence], 'B_block_length': [self.B_block_length], 'A_block_length': [self.A_block_length], 'blockiness_list': [self.blockiness_list], 'mean_blockiness': [self.mean_blockiness], 'perc_A_actual': [self.perc_A_actual], 'A_actual': [self.A_actual] } df = pd.DataFrame(data) df.to_csv(filename, index=False)
[docs] def pack_solvated_system(self, salt_concentration=0.0* unit.mole / unit.liter, residual_monomer=0.00): """ Pack a solvated system using Packmol functions, and the OpenFF Packmol wrapper. This method uses Packmol to build a solvated system by packing molecules into a simulation box. It considers the salt concentration and residual monomer concentration to determine the quantity of each molecule type required. Parameters ---------- salt_concentration : openff.units.Quantity, optional The desired salt concentration in the simulation box. Default is 0.0 M. residual_monomer : float, optional The desired residual monomer concentration in the simulation box. Default is 0.00. Returns ------- openff.toolkit.topology.topology.Topology An Interchange object representing the packed solvated system. Notes ----- This function uses the OpenFF Interchange Packmol wrapper to pack the molecules into the simulation box. It removes any molecules with zero copies before packing, to avoid packmol errors. Assigns % residual monomer value to ensemble under self.residual_monomer_actual """ from openff.interchange.components._packmol import pack_box, UNIT_CUBE from swiftpol.build import calculate_box_components molecules, number_of_copies, topology, box_vectors, residual_monomer_actual = calculate_box_components( self.chains, self.monomers, self.sequence, salt_concentration, residual_monomer ) molecules = [molecules[i] for i in range(len(number_of_copies)) if number_of_copies[i] != 0] number_of_copies = [num for num in number_of_copies if num != 0] self.residual_monomer_actual = residual_monomer_actual if topology.n_molecules == 1: return pack_box(molecules=molecules, number_of_copies=number_of_copies, solute=topology, box_vectors=box_vectors, box_shape=UNIT_CUBE, center_solute = 'BRICK', working_directory = '.',) else: return pack_box(molecules = molecules + self.chains, number_of_copies = number_of_copies+[1 for i in range(len(self.chains))], box_vectors = box_vectors, tolerance = 1*unit.angstrom)
[docs] def generate_polyply_files(self, residual_monomer=0.00, residual_oligomer=0.00, ): """ Generate input files for Polyply from the system. This method generates the input files required for Polyply (https://github.com/marrink-lab/polyply_1.0/) from the system. Parameters ---------- residual_monomer : float, optional The desired residual monomer concentration in the simulation box. Default is 0.00. residual_oligomer : float, optional The desired residual oligomer concentration in the simulation box. Default is 0.00. Returns ------- tuple A tuple containing the paths to the .gro, .top, and .pdb files generated for Polyply and GROMACS insert-molecules input. Notes ----- This function uses OpenFF Interchange to generate the input files for Polyply. Raises ------ UserWarning If partial charges are not assigned to the system, processing large systems may raise errors from OpenFF-Interchange. """ from swiftpol.build import calculate_box_components from openff.interchange import Interchange from openff.toolkit import ForceField import warnings box_vectors = calculate_box_components(self.chains, self.monomers, self.sequence, salt_concentration = 0.0* unit.mole / unit.liter, residual_monomer = 0.00)[3] molecules, number_of_copies, residual_monomer_actual, residual_oligomer_actual = self.calculate_residuals(residual_monomer, residual_oligomer) mol_pdb_files_dest = [] for i in molecules: string_i = str(molecules.index(i)) + '.pdb' mol_pdb_files_dest.append(string_i) i.generate_conformers(n_conformers=1) i.to_file(string_i, file_format='pdb') self.residual_monomer_actual = residual_monomer_actual self.residual_oligomer_actual = residual_oligomer_actual topology = Topology.from_molecules(self.chains+[molecules[i] for i in range(len(number_of_copies)) if number_of_copies[i] != 0]) if self.chains[0].partial_charges is None: warnings.warn('Partial charges may not be assigned to the system. Processing large systems may raise errors from OpenFF-Interchange', UserWarning) interchange = Interchange.from_smirnoff( topology=topology, force_field=ForceField("openff-2.2.0.offxml"), box=box_vectors ) else: interchange = Interchange.from_smirnoff( topology=topology, force_field=ForceField("openff-2.2.0.offxml"), charge_from_molecules=list(set([i for i in self.chains])), box=box_vectors ) interchange.to_gromacs('swiftpol_output') return_tuple = ('swiftpol_output.gro', 'swiftpol_output.top') for pdb_file in mol_pdb_files_dest: return_tuple += (pdb_file,) print(f'Polyply input files generated! Saved at {return_tuple}') return return_tuple
[docs] def calculate_residuals(self, residual_monomer = 0, residual_oligomer = 0, return_rdkit = False): """ Generate residual monomer and oligomer molecules, and molecule counts. Parameters ---------- residual_monomer : float The desired residual monomer concentration in the simulation box. Default is 0.00. residual_oligomer : float The desired residual oligomer concentration in the simulation box. Default is 0.00. return_rdkit : bool, optional If True, return RDKit molecule objects instead of OpenFF Molecule objects. Default is False. Returns ------- tuple A tuple containing the following elements: - A list of OpenFF Molecule objects representing the residual monomer and oligomer molecules. - A list of molecule counts, corresponding with the molecule list - Actual value for % residual monomer - Actual value for % residual oligomer Notes ----- EXPERIMENTAL CAPABILITY. Proceed with caution This function works independently of, and is unrelated to, the build.calculate_box_components function. Raises ------ UserWarning - If the residual monomer concentration is close to or lower than the weight of a single monomer. """ from functools import reduce from swiftpol import build import numpy as np from openff.toolkit import Topology, Molecule from rdkit import Chem import warnings topology = Topology.from_molecules(self.chains) rolling_mass=0 monomers = self.monomers sequence = self.sequence for m in topology.molecules: rolling_mass += sum(atom.mass for atom in m.atoms) if rolling_mass.magnitude < 1000: warnings.warn('Residual monomer/oligomer calculation may not be appropriate for small systems, as the residual mass can be close to or lower than the weight of a single monomer. Please check the output by querying the value of residual_monomer_actual') number_of_copies = [] # residual to add monomer_to_add = (rolling_mass.magnitude/100 - residual_monomer - residual_oligomer) * residual_monomer oligo_to_add = (rolling_mass.magnitude/100 - residual_monomer - residual_oligomer) * residual_oligomer if 'A' in sequence and 'B' in sequence: A_rd = Chem.MolFromSmiles(monomers[0]) A_rd = Chem.AddHs(A_rd) info = Chem.AtomPDBResidueInfo() info.SetResidueName('A' + str(1)) info.SetResidueNumber(1) [atom.SetMonomerInfo(info) for atom in A_rd.GetAtoms()] A = Molecule.from_rdkit(A_rd) A.name = 'A_residual' A_mass = sum([atom.mass for atom in A.atoms]) B_rd = Chem.MolFromSmiles(monomers[1]) B_rd = Chem.AddHs(B_rd) info = Chem.AtomPDBResidueInfo() info.SetResidueName('B' + str(1)) info.SetResidueNumber(1) [atom.SetMonomerInfo(info) for atom in B_rd.GetAtoms()] B = Molecule.from_rdkit(B_rd) B_mass = sum([atom.mass for atom in B.atoms]) B.name = 'B_residual' mon_count = 0 for r in range(1,100): if (r * A_mass.magnitude) + (r * B_mass.magnitude) >= monomer_to_add: mon_count = r break A_to_add = mon_count B_to_add = mon_count elif 'A' in sequence and 'B' not in sequence: A_rd = Chem.MolFromSmiles(monomers[0]) A_rd = Chem.AddHs(A_rd) info = Chem.AtomPDBResidueInfo() info.SetResidueName('A' + str(1)) info.SetResidueNumber(1) [atom.SetMonomerInfo(info) for atom in A_rd.GetAtoms()] A = Molecule.from_rdkit(A_rd) A_mass = sum([atom.mass for atom in A.atoms]) A.name = 'A_residual' B = Molecule.from_smiles('C') B.name = 'B_residual' mon_count = 0 for r in range(1,100): if (r * A_mass.magnitude) >= monomer_to_add: mon_count = r break A_to_add = mon_count B_to_add = 0 oligomers = [] if 'A' in sequence and 'B' in sequence: for i in range(1000): sigma = np.sqrt(np.log(1.05)) mu = np.log(self.length_target * 0.1) - 0.5 * sigma**2 chain_lengths = np.random.lognormal(mu, sigma, size=1) chain_lengths = np.round(chain_lengths).astype(int) oligo_seq = reduce(lambda x, y: x + y, np.random.choice(['A', 'B'], size=(int(chain_lengths)), p=[self.A_target/100,1-(self.A_target/100)])) monomer_list = [mono+'[I]' for mono in self.monomers] oligomer_rd = build.build_polymer(oligo_seq, monomer_list=monomer_list, reaction=AllChem.ReactionFromSmarts(self.reaction), chain_num=len(self.chains)+1+i) oligomer_rd = Chem.AddHs(oligomer_rd) oligomer = Molecule.from_rdkit(oligomer_rd) oligomer.name = 'oligo' + str(len(self.chains)+1+i) oligo_mass = 0 oligomers_new = oligomers + [oligomer] for i in oligomers_new: oligo_mass += sum(atom.mass for atom in i.atoms) if oligo_mass.magnitude <= oligo_to_add: oligomers.append(oligomer) else: break monomer_mass = (A_to_add * A_mass.magnitude) + (B_to_add * B_mass.magnitude) oligomer_mass = 0 * unit.dalton for i in oligomers: oligomer_mass += sum(atom.mass for atom in i.atoms) residual_monomer_actual = (monomer_mass / (rolling_mass.magnitude + monomer_mass + oligomer_mass.magnitude)) * 100 residual_oligomer_actual = (oligomer_mass.magnitude / (rolling_mass.magnitude + monomer_mass + oligomer_mass.magnitude)) * 100 molecules = [A, B] + oligomers number_of_copies = [A_to_add, B_to_add] + [1 for c in range(len(oligomers))] elif 'A' in sequence and 'B' not in sequence: for i in range(1000): sigma = np.sqrt(np.log(1.05)) mu = np.log(self.length_target * 0.1) - 0.5 * sigma**2 chain_lengths = np.random.lognormal(mu, sigma, size=1) chain_lengths = np.round(chain_lengths).astype(int) oligo_seq = int(50 * 0.1) * 'A' monomer_list = [mono+'[I]' for mono in self.monomers] oligomer_rd = build.build_polymer(oligo_seq, monomer_list=monomer_list, reaction=AllChem.ReactionFromSmarts(self.reaction), chain_num=len(self.chains)+1+i) oligomer_rd = Chem.AddHs(oligomer_rd) oligomer = Molecule.from_rdkit(oligomer_rd) oligomer.name = 'oligo' + str(len(self.chains)+1+i) oligo_mass = 0 oligomers_new = oligomers + [oligomer] for i in oligomers_new: oligo_mass += sum(atom.mass for atom in i.atoms) if oligo_mass.magnitude <= oligo_to_add: oligomers.append(oligomer) else: break B_to_add = 0 monomer_mass = (A_to_add * A_mass.magnitude) + (B_to_add * 0) oligomer_mass = 0 * unit.dalton for i in oligomers: oligomer_mass += sum(atom.mass for atom in i.atoms) residual_monomer_actual = (monomer_mass / (rolling_mass.magnitude + monomer_mass + oligomer_mass.magnitude)) * 100 residual_oligomer_actual = (oligomer_mass.magnitude / (rolling_mass.magnitude + monomer_mass + oligomer_mass.magnitude)) * 100 molecules = [A, B] + oligomers number_of_copies = [A_to_add, B_to_add] + [1 for c in range(len(oligomers))] if return_rdkit: molecules = [mol.to_rdkit() for mol in molecules] return molecules, number_of_copies, residual_monomer_actual, residual_oligomer_actual else: for i in molecules: i.generate_unique_atom_names() #required for polyply output return molecules, number_of_copies, residual_monomer_actual, residual_oligomer_actual
[docs] class polymer_system_from_PDI: """ Build a polymer system based on a target polydispersity index (PDI). This class generates a polymer system with a specified number of chains, target chain length, and target PDI. It supports the generation of copolymers, stereoisomers, and blockiness control. The system is built using OpenFF and RDKit toolkits, with optional support for OpenEye toolkits if installed. **Note**: This capability is under development and may change significantly in the next version of SwiftPol. Proceed with caution. Parameters ---------- monomer_list : list of str A list of monomer SMILES strings to be used in the polymer system. reaction : str A SMARTS string defining the polymerization reaction. length_target : int The target average chain length for the polymer system. num_chains : int The number of polymer chains to generate. PDI_target : float The target polydispersity index (PDI) for the polymer system. stereoisomerism_input : tuple, optional A tuple containing the stereoisomer monomer, the fraction of stereoisomers to introduce, and the SMILES string of the stereoisomer. Default is None. terminals : str, optional The type of terminal groups to use for the polymer chains. Default is 'hydrogen'. perc_A_target : float, optional The target percentage of monomer 'A' in the copolymer. Default is 100. blockiness_target : list, optional A list containing the target blockiness value and the monomer to control blockiness for. Default is [1.0, 'A']. copolymer : bool, optional Whether to generate a copolymer. Default is False. acceptance : float, optional The acceptance margin (in percentage) for the target blockiness and monomer percentage. Default is 10. Attributes ---------- chains : list of openff.toolkit.topology.Molecule A list of OpenFF Molecule objects representing the polymer chains. chain_rdkit : list of rdkit.Chem.Mol A list of RDKit Molecule objects representing the polymer chains. sequence : str The sequence of monomers in the polymer system. mol_weight_average : float The average molecular weight of the polymer chains. PDI : float The polydispersity index of the polymer system. Mn : float The number-average molecular weight of the polymer system. Mw : float The weight-average molecular weight of the polymer system. num_chains : int The number of polymer chains in the system. length_average : float The average chain length of the polymer system. lengths : list of int A list of chain lengths for the polymer chains. min_length : int The minimum chain length in the polymer system. max_length : int The maximum chain length in the polymer system. A_actual : float or None The actual percentage of monomer 'A' in the copolymer. None if not a copolymer. perc_A_actual : list of float or None A list of the actual percentages of monomer 'A' in each chain. None if not a copolymer. B_block_length : float or None The average block length of monomer 'B'. None if not a copolymer. A_block_length : float or None The average block length of monomer 'A'. None if not a copolymer. blockiness_list : list of float or None A list of blockiness values for each chain. None if not a copolymer. mean_blockiness : float or None The mean blockiness value for the polymer system. None if not a copolymer. Notes ----- - If OpenEye toolkits are not installed, RDKit will be used for conformer generation. - The system is built using a log-normal distribution of chain lengths to achieve the target PDI. - For small systems, PDI may not be close to the target value. Examples -------- >>> from swiftpol.build import polymer_system_from_PDI >>> monomer_list=['OC(=O)COI'] >>> reaction = '[C:1][O:2][H:3].[I:4][O:5][C:6]>>[C:1][O:2][C:6].[H:3][O:5][I:4]' >>> system = polymer_system_from_PDI( ... monomer_list=monomer_list, ... reaction=reaction, ... length_target=50, ... num_chains=50, ... PDI_target=1.7, ... copolymer=False, ... ) >>> print(system) SwiftPol ensemble of size 50, average chain length = 10.6-mers, PDI = 1.7006109664210667 """ try: from openeye import oechem except: import warnings warnings.warn("OpenEye is not installed. You will not be able to use OpenEye Toolkits for conformer generation.") from openff.toolkit.utils.toolkits import RDKitToolkitWrapper, OpenEyeToolkitWrapper from functools import reduce from statistics import mean from rdkit.Chem.Descriptors import ExactMolWt from openff.interchange import Interchange from openff.interchange.components._packmol import UNIT_CUBE, pack_box from swiftpol.build import build_polymer, PDI, blockiness_gen, calculate_box_components, introduce_stereoisomers from openff.units import unit from rdkit.Chem import AllChem
[docs] def __init__(self, monomer_list, reaction, length_target, num_chains, PDI_target, stereoisomerism_input=None, terminals='hydrogen', perc_A_target=100, blockiness_target=[1.0, 'A'], copolymer=False, acceptance = 10 ): self.length_target = length_target self.terminals = terminals perc_A_actual = [] if copolymer==True: self.blockiness_target = blockiness_target[0] self.A_target = perc_A_target def generate_chain_lengths_from_PDI(Mn_target, PDI_target, num_chains, clip_extremes=True): sigma = np.sqrt(np.log(PDI_target*2.5)) mu = np.log(Mn_target) - 0.5 * sigma**2 lengths = np.random.lognormal(mu, sigma, size=num_chains) if clip_extremes: min_length = 1 max_length = 5 * Mn_target lengths = np.clip(lengths, min_length, max_length) scale_factor = Mn_target / np.mean(lengths) lengths *= scale_factor return list(lengths) def spec(sequence): #Define limits of A percentage and blockiness from input acceptance_dec = acceptance/100 actual_A = (sequence.count('A')/len(sequence))*100 blockiness = blockiness_gen(sequence, blockiness_target[1])[0] return actual_A > perc_A_target*(1-acceptance_dec) and actual_A < perc_A_target*(1+acceptance_dec) and blockiness>blockiness_target[0]*(1-acceptance_dec) and blockiness<blockiness_target[0]*(1+acceptance_dec) blockiness_list = [] out_of_spec = 0 BBL = [] ABL = [] chains = [] chains_rdkit = [] lengths = [] self.monomers = [mono.replace("[I]", "") for mono in monomer_list] self.reaction = reaction reaction = AllChem.ReactionFromSmarts(reaction) if stereoisomerism_input is not None: stereo_monomer, instance, new_smiles = stereoisomerism_input monomer_list.append(new_smiles) # Calculate range of chain lengths needed to build around a particular PDI sigma = np.sqrt(np.log(PDI_target)) mu = np.log(length_target) - 0.5 * sigma**2 chain_lengths = np.random.lognormal(mu, sigma, size=num_chains) chain_lengths = np.round(chain_lengths).astype(int) #First round of building - copolymer #First round of building - copolymer if copolymer==True: n=0 length_list = generate_chain_lengths_from_PDI(length_target, PDI_target, num_chains, clip_extremes=True) for l in length_list: while True: sequence = reduce(lambda x, y: x + y, np.random.choice(['A', 'B'], size=(int(l),), p=[perc_A_target/100,1-(perc_A_target/100)])) blockiness = blockiness_gen(sequence, blockiness_target[1])[0] if spec(sequence): break # Exit the loop if the sequence is valid blockiness = blockiness_gen(sequence, blockiness_target[1])[0] if spec(sequence)==True: if stereoisomerism_input is not None: sequence_stereo = introduce_stereoisomers(stereo_monomer, instance, sequence) pol = build_polymer(sequence=sequence_stereo, monomer_list = monomer_list, reaction = reaction, terminal=terminals, chain_num=n+1) else: pol = build_polymer(sequence=sequence, monomer_list = monomer_list, reaction = reaction, terminal=terminals, chain_num=n+1) lengths.append(int(l)) chains_rdkit.append(pol) chain = Molecule.from_rdkit(pol) chains.append(chain) n+=1 perc_A_actual.append((sequence.count('A')/len(sequence))*100) blockiness_list.append(blockiness) BBL.append(blockiness_gen(sequence, blockiness_target[1])[1]) ABL.append(blockiness_gen(sequence, blockiness_target[1])[2]) self.B_block_length = mean(BBL) self.A_block_length = mean(ABL) self.blockiness_list = blockiness_list self.mean_blockiness = mean(blockiness_list) self.perc_A_actual = perc_A_actual self.A_actual = mean(perc_A_actual) else: n=0 for l in chain_lengths: sequence = reduce(lambda x, y: x + y, np.random.choice(['A', 'B'], size=(int(l),), p=[perc_A_target/100,1-(perc_A_target/100)])) if stereoisomerism_input is not None: sequence_stereo = introduce_stereoisomers(stereo_monomer, instance, sequence) pol = build_polymer(sequence=sequence_stereo, monomer_list = monomer_list, reaction = reaction, terminal=terminals, chain_num=n+1) else: pol = build_polymer(sequence=sequence, monomer_list = monomer_list, reaction = reaction, terminal=terminals, chain_num=n+1) lengths.append(int(l)) chains_rdkit.append(pol) chain = Molecule.from_rdkit(pol) chains.append(chain) perc_A_actual.append((sequence.count('A')/len(sequence))*100) n+=1 self.B_block_length = None self.A_block_length = None self.blockiness_list = None self.mean_blockiness = None self.perc_A_actual = None self.A_actual = None self.sequence = sequence self.chains = chains for i in range(len(self.chains)): self.chains[i].name = 'chain' + str(i+1) self.chain_rdkit = chains_rdkit self.mol_weight_average = round(mean([ExactMolWt(c) for c in chains_rdkit]),2) self.PDI, self.Mn, self.Mw = PDI(chains_rdkit) self.num_chains = len(chains) self.A_actual = mean(perc_A_actual) self.perc_A_actual = perc_A_actual self.length_average = mean(lengths) self.lengths = lengths self.min_length = min(lengths) self.max_length = max(lengths) print('System built!, size =', self.num_chains)
def __repr__(self): description = (f"SwiftPol ensemble of size {self.num_chains}, " f"average chain length = {self.length_average}-mers, PDI = {self.PDI}") return description
[docs] def generate_conformers(self): """ Generate conformers for each polymer chain in the system. This method uses the OpenFF toolkit OpenEye Wrapper to generate conformers for each polymer chain in the system. It first checks if the OpenEye toolkit is licensed and available. If it is, it uses the OpenEyeToolkitWrapper to generate conformers. Otherwise, it falls back to using the RDKitToolkitWrapper. Each chain is processed to generate a single conformer, and unique atom names are assigned to each chain. Raises ------ ImportError If neither RDKit nor OpenEye toolkits are available. """ from openff.toolkit.utils.toolkits import RDKitToolkitWrapper, OpenEyeToolkitWrapper # Generate conformers using OpenFF toolkit wrapper for chain in self.chains: num = self.chains.index(chain) object = RDKitToolkitWrapper() if oechem_imported: if oechem.OEChemIsLicensed(): object = OpenEyeToolkitWrapper() object.generate_conformers(molecule=chain, n_conformers=1) chain.generate_unique_atom_names() self.chains[num] = chain
[docs] def charge_system(self, charge_scheme): """ Assign partial charges to each polymer chain in the system. This method uses one of AM1-BCC, Espaloma, or OpenFF NAGL to assign partial charges to each polymer chain in the system. It iterates over each chain in the `self.chains` list and assigns partial charges to the chain. Parameters ---------- charge_scheme : str The charge assignment scheme to use. Options are 'AM1_BCC', 'espaloma', or 'NAGL'. Raises ------ ImportError If the selected toolkit is not available. """ from swiftpol.parameterize import charge_openff_polymer for chain in self.chains: chain.partial_charges = charge_openff_polymer(chain, charge_scheme)
[docs] def export_to_csv(self, filename): """ Export all the instances in polymer_system.__init__() into a pandas DataFrame and save it as a CSV file. Parameters ---------- filename : str The name of the CSV file to save the data. """ data = { 'monomers': [self.monomers], 'length_target': [self.length_target], 'terminals': [self.terminals], 'num_chains': [self.num_chains], 'mol_weight_average': [self.mol_weight_average], 'PDI': [self.PDI], 'Mn': [self.Mn], 'Mw': [self.Mw], 'sequence': [self.sequence], 'B_block_length': [self.B_block_length], 'A_block_length': [self.A_block_length], 'blockiness_list': [self.blockiness_list], 'mean_blockiness': [self.mean_blockiness], 'perc_A_actual': [self.perc_A_actual], 'A_actual': [self.A_actual] } df = pd.DataFrame(data) df.to_csv(filename, index=False)
[docs] def pack_solvated_system(self, salt_concentration=0.0, residual_monomer=0.00): """ Pack a solvated system using Packmol functions, and the OpenFF Packmol wrapper. This method uses Packmol to build a solvated system by packing molecules into a simulation box with water. It considers the salt concentration and residual monomer concentration to determine the quantity of each molecule type required. Parameters ---------- salt_concentration : openff.units.Quantity, optional The desired salt concentration in the simulation box. Default is 0.0 M. residual_monomer : float, optional The desired residual monomer concentration in the simulation box. Default is 0.00. Returns ------- openff.toolkit.topology.topology.Topology An Interchange object representing the packed solvated system. Notes ----- This function uses the OpenFF Interchange Packmol wrapper to pack the molecules into the simulation box. It removes any molecules with zero copies before packing, to avoid packmol errors. Assigns % residual monomer value to ensemble under self.residual_monomer_actual """ from openff.interchange.components._packmol import pack_box, UNIT_CUBE from swiftpol.build import calculate_box_components molecules, number_of_copies, topology, box_vectors, residual_monomer_actual = calculate_box_components( self.chains, self.monomers, self.sequence, salt_concentration, residual_monomer ) molecules = [molecules[i] for i in range(len(number_of_copies)) if number_of_copies[i] != 0] number_of_copies = [num for num in number_of_copies if num != 0] self.residual_monomer_actual = residual_monomer_actual if topology.n_molecules == 1: return pack_box(molecules=molecules, number_of_copies=number_of_copies, solute=topology, box_vectors=box_vectors, box_shape=UNIT_CUBE, center_solute = 'BRICK') else: return pack_box(molecules = molecules + self.chains, number_of_copies = number_of_copies+[1 for i in range(len(self.chains))], box_vectors = box_vectors, tolerance = 1*unit.angstrom)
[docs] def generate_polyply_files(self, residual_monomer=0.00, residual_oligomer=0.00): """ Generate input files for Polyply from the system. This method generates the input files required for Polyply (https://github.com/marrink-lab/polyply_1.0/) from the system. Parameters ---------- residual_monomer : float, optional The desired residual monomer concentration in the simulation box. Default is 0.00. residual_oligomer : float, optional The desired residual oligomer concentration in the simulation box. Default is 0.00. Returns ------- tuple A tuple containing the paths to the .gro, .top, and .pdb files generated for Polyply and GROMACS insert-molecules input. Notes ----- This function uses OpenFF Interchange to generate the input files for Polyply. Raises ------ UserWarning If partial charges are not assigned to the system, processing large systems may raise errors from OpenFF-Interchange. """ from swiftpol.build import calculate_box_components from openff.interchange import Interchange from openff.toolkit import ForceField import warnings box_vectors = calculate_box_components(self.chains, self.monomers, self.sequence, 0.0, 0.0)[3] molecules, number_of_copies, residual_monomer_actual, residual_oligomer_actual = self.calculate_residuals(residual_monomer, residual_oligomer) mol_pdb_files_dest = [] for i in molecules: string_i = str(molecules.index(i)) + '.pdb' mol_pdb_files_dest.append(string_i) if i.has_unique_atom_names == False: i.generate_unique_atom_names() i.generate_conformers(n_conformers=1) i.to_file(string_i, file_format='pdb') self.residual_monomer_actual = residual_monomer_actual self.residual_oligomer_actual = residual_oligomer_actual topology = Topology.from_molecules(self.chains+[molecules[i] for i in range(len(number_of_copies)) if number_of_copies[i] != 0]) if self.chains[0].partial_charges is None: warnings.warn('Partial charges may not be assigned to the system. Processing large systems may raise errors from OpenFF-Interchange', UserWarning) interchange = Interchange.from_smirnoff( topology=topology, force_field=ForceField("openff-2.2.0.offxml"), box=box_vectors ) else: interchange = Interchange.from_smirnoff( topology=topology, force_field=ForceField("openff-2.2.0.offxml"), charge_from_molecules=[i for i in self.chains], box=box_vectors ) interchange.to_gromacs('swiftpol_output') return_tuple = ('swiftpol_output.gro', 'swiftpol_output.top') for pdb_file in mol_pdb_files_dest: return_tuple += (pdb_file,) print(f'Polyply input files generated! Saved at {return_tuple}') return return_tuple
[docs] def calculate_residuals(self, residual_monomer = 0, residual_oligomer = 0): """ Generate residual monomer and oligomer molecules, and molecule counts. Parameters ---------- residual_monomer : float The desired residual monomer concentration in the simulation box. Default is 0.00. residual_oligomer : float The desired residual oligomer concentration in the simulation box. Default is 0.00. Returns ------- tuple A tuple containing the following elements: - A list of OpenFF Molecule objects representing the residual monomer and oligomer molecules. - A list of molecule counts, corresponding with the molecule list - Actual value for % residual monomer - Actual value for % residual oligomer Notes ----- EXPERIMENTAL CAPABILITY. Proceed with caution This function works independently of, and is unrelated to, the build.calculate_box_components function. Residual oligomers are polydisperse and are generated based on 10% of the target chain length used in ensemble initiation. Raises ------ UserWarning - If the residual monomer concentration is close to or lower than the weight of a single monomer. """ from functools import reduce from swiftpol import build import numpy as np from openff.toolkit import Topology, Molecule from rdkit import Chem import warnings topology = Topology.from_molecules(self.chains) rolling_mass=0 monomers = self.monomers sequence = self.sequence for m in topology.molecules: rolling_mass += sum(atom.mass for atom in m.atoms) if rolling_mass.magnitude < 1000: warnings.warn('Residual monomer/oligomer calculation may not be accurate for small systems, as the residual mass can be close to or lower than the weight of a single monomer. Please check the output by querying the value of residual_monomer_actual') number_of_copies = [] # residual to add monomer_to_add = (rolling_mass.magnitude/100 - residual_monomer - residual_oligomer) * residual_monomer oligo_to_add = (rolling_mass.magnitude/100 - residual_monomer - residual_oligomer) * residual_oligomer if 'A' in sequence and 'B' in sequence: A_rd = Chem.MolFromSmiles(monomers[0]) A_rd = Chem.AddHs(A_rd) info = Chem.AtomPDBResidueInfo() info.SetResidueName('A' + str(1)) info.SetResidueNumber(1) [atom.SetMonomerInfo(info) for atom in A_rd.GetAtoms()] A = Molecule.from_rdkit(A_rd) A.name = 'A_residual' A_mass = sum([atom.mass for atom in A.atoms]) B_rd = Chem.MolFromSmiles(monomers[1]) B_rd = Chem.AddHs(B_rd) info = Chem.AtomPDBResidueInfo() info.SetResidueName('B' + str(1)) info.SetResidueNumber(1) [atom.SetMonomerInfo(info) for atom in B_rd.GetAtoms()] B = Molecule.from_rdkit(B_rd) B_mass = sum([atom.mass for atom in B.atoms]) B.name = 'B_residual' for r in range(1,100): if (r * A_mass.magnitude) + (r * B_mass.magnitude) >= monomer_to_add: mon_count = r break A_to_add = mon_count B_to_add = mon_count elif 'A' in sequence and 'B' not in sequence: A_rd = Chem.MolFromSmiles(monomers[0]) A_rd = Chem.AddHs(A_rd) info = Chem.AtomPDBResidueInfo() info.SetResidueName('A' + str(1)) info.SetResidueNumber(1) [atom.SetMonomerInfo(info) for atom in A_rd.GetAtoms()] A = Molecule.from_rdkit(A_rd) A_mass = sum([atom.mass for atom in A.atoms]) A.name = 'A_residual' B = Molecule.from_smiles('C') B.name = 'B_residual' for r in range(1,100): if (r * A_mass.magnitude) >= monomer_to_add: mon_count = r break A_to_add = mon_count B_to_add = 0 oligomers = [] if 'A' in sequence and 'B' in sequence: for i in range(1000): oligo_seq = reduce(lambda x, y: x + y, np.random.choice(['A', 'B'], size=(int(self.length_target * 0.1)), p=[self.A_target/100,1-(self.A_target/100)])) monomer_list = [mono+'[I]' for mono in self.monomers] oligomer_rd = build.build_polymer(oligo_seq, monomer_list=monomer_list, reaction=AllChem.ReactionFromSmarts(self.reaction)) oligomer_rd = Chem.AddHs(oligomer_rd) info = Chem.AtomPDBResidueInfo() info.SetResidueName('O' + str(i+1)) info.SetResidueNumber(1) [atom.SetMonomerInfo(info) for atom in oligomer_rd.GetAtoms()] oligomer = Molecule.from_rdkit(oligomer_rd) oligo_mass = 0 oligomers_new = oligomers + [oligomer] for i in oligomers_new: oligo_mass += sum(atom.mass for atom in i.atoms) if oligo_mass.magnitude <= oligo_to_add: oligomers.append(oligomer) else: break monomer_mass = (A_to_add * A_mass.magnitude) + (B_to_add * B_mass.magnitude) oligomer_mass = 0 * unit.dalton for i in oligomers: oligomer_mass += sum(atom.mass for atom in i.atoms) residual_monomer_actual = (monomer_mass / (rolling_mass.magnitude + monomer_mass + oligomer_mass.magnitude)) * 100 residual_oligomer_actual = (oligomer_mass.magnitude / (rolling_mass.magnitude + monomer_mass + oligomer_mass.magnitude)) * 100 molecules = [A, B] + oligomers number_of_copies = [A_to_add, B_to_add] + [1 for c in range(len(oligomers))] elif 'A' in sequence and 'B' not in sequence: for i in range(1000): oligo_seq = reduce(lambda x, y: x + y, np.random.choice(['A'], size=(int(self.length_target * 0.1)), p=[self.A_target/100,1-(self.A_target/100)])) monomer_list = [mono+'[I]' for mono in self.monomers] oligomer_rd = build.build_polymer(oligo_seq, monomer_list=monomer_list, reaction=AllChem.ReactionFromSmarts(self.reaction)) oligomer_rd = Chem.AddHs(oligomer_rd) info = Chem.AtomPDBResidueInfo() info.SetResidueName('O' + str(i+1)) info.SetResidueNumber(1) [atom.SetMonomerInfo(info) for atom in oligomer_rd.GetAtoms()] oligomer = Molecule.from_rdkit(oligomer_rd) oligomer.name = 'oligo' + str(i+1) oligo_mass = 0 oligomers_new = oligomers + [oligomer] for i in oligomers_new: oligo_mass += sum(atom.mass for atom in i.atoms) if oligo_mass.magnitude <= oligo_to_add: oligomers.append(oligomer) else: break monomer_mass = (A_to_add * A_mass.magnitude) + (B_to_add * B_mass.magnitude) oligomer_mass = 0 * unit.dalton for i in oligomers: oligomer_mass += sum(atom.mass for atom in i.atoms) residual_monomer_actual = (monomer_mass / (rolling_mass.magnitude + monomer_mass + oligomer_mass.magnitude)) * 100 residual_oligomer_actual = (oligomer_mass.magnitude / (rolling_mass.magnitude + monomer_mass + oligomer_mass.magnitude)) * 100 molecules = [A, B] + oligomers number_of_copies = [A_to_add, B_to_add] + [1 for c in range(len(oligomers))] return molecules, number_of_copies, residual_monomer_actual, residual_oligomer_actual