Motivation

I often have to modify a given molecule to introduce a set of modification to make congeneric series. AllChem.ConstrainedEmbed in RDKit could provide such function. See below blog posts for an example:

However, AllChem.ConstrinedEmbed uses MCS algorithm, and the MCS sometimes did not yielded the core that I desired, which resulted in a completely wrong alignment between the parent and the newly modified molecule.

So, I wrote a small function that takes the SMARTS pattern, and explicitly takes the coordinates then embed the rest of the coordinates.

from io import BytesIO
import pandas as pd
import numpy as np
from rdkit.Chem import PandasTools

from rdkit import Chem
from rdkit.Chem import AllChem
from rdkit.Chem import DataStructs
from rdkit.Chem import rdMolDescriptors
from rdkit.Chem import rdRGroupDecomposition

from rdkit.Chem.Draw import IPythonConsole #Needed to show molecules
from rdkit.Chem import Draw
from rdkit.Chem import rdDepictor
from rdkit.Chem.Draw import rdMolDraw2D
from rdkit.Chem.Draw.MolDrawing import MolDrawing, DrawingOptions #Only needed if modifying defaults

DrawingOptions.bondLineWidth=1.8
IPythonConsole.ipython_useSVG=True
from rdkit import RDLogger
RDLogger.DisableLog('rdApp.warning')
import rdkit
import py3Dmol
print(rdkit.__version__)

2020.03.2

Example

Let's use Imatinib for the parent molecule as an example.

imatinib = Chem.MolFromMolFile('files/STI.sdf')
imatinib_2d = Chem.RemoveHs(imatinib)
AllChem.Compute2DCoords(imatinib_2d)
imatinib_2d
N N N NH HN O N N

And suppose we want to replace the piperazine to a phenyl.

mol_new = Chem.MolFromSmiles('Cc1ccc(NC(=O)c2ccc(c3ccccc3)cc2)cc1Nc1nccc(-c2cccnc2)n1')
mol_new
HN O NH N N N

SMARTS Pattern 

smarts = '[#6]1:[#6]:[#7]:[#6]:[#6](:[#6]:1)-[#6]1:[#7]:[#6](:[#7]:[#6]:[#6]:1)-[#7]-[#6]1:[#6]:[#6](:[#6]:[#6]:[#6]:1-[#6])-[#7]-[#6](-[#6]1:[#6]:[#6]:[#6](:[#6]:[#6]:1))=[#8]'
p = Chem.MolFromSmarts(smarts)
p
N N N N N O 
imatinib_2d.GetSubstructMatch(p)
imatinib_2d
N N N NH HN O N N 
mol_new.GetSubstructMatch(p)
mol_new
HN O NH N N N

Build the new conformer 

from rdkit.Chem import rdFMCS
from rdkit.Chem import rdDistGeom
from rdkit.Chem.rdForceFieldHelpers import UFFGetMoleculeForceField

def build(ref, m, smarts):
    mol = Chem.AddHs(m)

    p = Chem.MolFromSmarts(smarts)
    match1 = ref.GetSubstructMatch(p)
    match2 = mol.GetSubstructMatch(p)

    coordMap = {}
    coreConf = ref.GetConformer(0)
    for i, idxI in enumerate(match2):
        corePtI = coreConf.GetAtomPosition(match1[i])
        coordMap[idxI] = corePtI

    algMap = [(j, match1[i]) for i, j in enumerate(match2)]
    tdist = 0.25
    
    confid = AllChem.EmbedMolecule(mol, coordMap=coordMap)
    rms = AllChem.AlignMol(mol, ref, atomMap=algMap)
    ff = UFFGetMoleculeForceField(mol, confId=0)
    conf = ref.GetConformer()
    for i in range(len(match1)):
        p = conf.GetAtomPosition(match1[i])
        pIdx = ff.AddExtraPoint(p.x, p.y, p.z, fixed=True) - 1
        ff.AddDistanceConstraint(pIdx, match2[i], 0, tdist, 100.)
    ff.Initialize()
    n = 4

    more = ff.Minimize(energyTol=1e-4, forceTol=1e-3)
    while more and n:
        more = ff.Minimize(energyTol=1e-4, forceTol=1e-3)
        n -= 1
    # realign
    rms = AllChem.AlignMol(mol, ref, atomMap=algMap)
    return mol

mol_new = build(imatinib, mol_new, smarts)

You can see the new molecule have very similar coordinates except the newly added phenyl ring in 3D.

viewer = py3Dmol.view(width=300, height=300)
viewer.addModel(Chem.MolToMolBlock(imatinib), 'mol')
viewer.addModel(Chem.MolToMolBlock(mol_new), 'mol')
viewer.setStyle({"stick":{}})
viewer.zoomTo()
viewer.show()

You appear to be running in JupyterLab (or JavaScript failed to load for some other reason). You need to install the 3dmol extension:
jupyter labextension install jupyterlab_3dmol

Note you can perform a fragment grafting instead of embedding. See https://pschmidtke.github.io/blog/rdkit/3d-editor/2021/01/23/grafting-fragments.html