sdf2table.py

import argparse
import re
import base64
from io import BytesIO

import cairosvg
from rdkit import Chem
from rdkit.Chem import AllChem
from rdkit.Chem.Draw import rdMolDraw2D
from rdkit.Chem.Draw.MolDrawing import MolDrawing, DrawingOptions #Only needed if modifying defaults
DrawingOptions.atomLabelFontSize = 55
DrawingOptions.dotsPerAngstrom = 100
DrawingOptions.bondLineWidth = 2.0

def sdf_to_html(mols, fields, use_js, add_smiles=False):
    atoms_prop = re.compile(r'atom\.[a-z]?prop\.')

    rows = []
    nl = "\n"
    for i, m in enumerate(mols):
        _m = Chem.RemoveHs(m)
        AllChem.Compute2DCoords(_m)

        if i == 0:
            # print header
            keys = m.GetPropsAsDict().keys()
            # add 'mol' at the front to print molecular structure first
            # also remove atom.prop.XXX
            keys = ['mol'] + [k for k in keys if not atoms_prop.match(k)]

            if fields != 'all':
                keys = [k for k in keys if k == 'mol' or k in fields.split(',')]

            cols = []
            props = m.GetPropsAsDict()
            # always print entry number and mol name
            cols.append(f"<th>No</th>")
            cols.append(f"<th>Name</th>")
            if add_smiles:
                cols.append(f"<th>SMILES</th>")

            for k in keys:
                cols.append(f'<th>{k}</th>')
            row = f"""<thead><tr>{nl.join(cols)}</tr></thead>"""
            rows.append(row)

        cols = []
        # always print entry number and mol name
        cols.append(f"<td>{i+1}</td>")
        cols.append(f"<td>{m.GetProp('_Name')}</td>")
        if add_smiles:
            cols.append(f"<td>{Chem.MolToSmiles(m)}</td>")

        props = m.GetPropsAsDict()
        for k in keys:
            if k == 'mol':
                drawer = rdMolDraw2D.MolDraw2DSVG(300, 200)
                drawer.DrawMolecule(_m)
                drawer.FinishDrawing()
                svg = drawer.GetDrawingText()
                cols.append(f"""<td>{svg}</td>""")
            else:
                cols.append(f'<td>{props.get(k, "")}</td>')

        row = f"""<tr>{nl.join(cols)}</tr>"""
        rows.append(row)

    header = ""
    footer = ""
    if use_js:
        header = """
<link rel="stylesheet" type="text/css" href="https://cdn.datatables.net/1.10.21/css/jquery.dataTables.css">
<script src="https://code.jquery.com/jquery-3.5.1.slim.min.js"   integrity="sha256-4+XzXVhsDmqanXGHaHvgh1gMQKX40OUvDEBTu8JcmNs="   crossorigin="anonymous"></script>
<script type="text/javascript" charset="utf8" src="https://cdn.datatables.net/1.10.21/js/jquery.dataTables.js"></script>
"""
        footer = """
<script>
$(document).ready( function () {
  $('#table').DataTable({
    paging: false
  });
} );
</script>
"""


    table = f"""
{header}
<table id="table">{nl.join(rows)}</table>
{footer}
"""
    return table

def main():
    parser = argparse.ArgumentParser(description='Process SDF file into a HTML table')
    parser.add_argument('sdfile', help='SDF file')
    parser.add_argument('--fields', default="all", help='fields to print')
    parser.add_argument('--smiles', action="store_true", help='add SMILES field')
    parser.add_argument('--js', default=True, help='fields to print')
    args = parser.parse_args()

    sdfile = args.sdfile
    mols = Chem.SDMolSupplier(sdfile, removeHs=False)
    html = sdf_to_html(mols, args.fields, args.js, args.smiles)
    print(html)

if __name__ == "__main__":
    main()