print 'Hello World!'

1+4

from rdkit.Chem import AllChem as Chem
from rdkit.Chem.Draw import IPythonConsole
from rdkit.Chem import Descriptors
from rdkit import DataStructs

smi = 'CCCc1nn(C)c2C(=O)NC(=Nc12)c3cc(ccc3OCC)S(=O)(=O)N4CCN(C)CC4' #sildenafil
m = Chem.MolFromSmiles(smi)

m

Descriptors.MolWt(m)

Descriptors.TPSA(m)

Descriptors.RingCount(m)

Chem.MolToSmiles(m, True)

Chem.MolToInchi(m)

print Chem.MolToMolBlock(m)

Chem.Compute2DCoords(m)

print Chem.MolToMolBlock(m)

fp = Chem.GetMorganFingerprintAsBitVect(m,2,nBits=2048)

fp.ToBitString()

fp.GetNumOnBits()

fp.GetNumBits()

smi2 = 'CCCc1nc(C)c2C(=O)N=C(Nn12)c3cc(ccc3OCC)S(=O)(=O)N4CCN(CC)CC4' #vardenafil
m2 = Chem.MolFromSmiles(smi2)

fp2 = Chem.GetMorganFingerprintAsBitVect(m2, 2, nBits=2048)

m2

m

DataStructs.TanimotoSimilarity(fp,fp2)

from rdkit.Chem.Draw import SimilarityMaps
SimilarityMaps.GetSimilarityMapForFingerprint(m2, m, SimilarityMaps.GetMorganFingerprint)

sma = 'C1C[!#1!#6]1' #oxirane or aziridine

from IPython.display import Image
from urllib import quote_plus

Image(url='http://www.smartsview.de/smartsview/auto/png/1/dynamic/{0}'.format(quote_plus(sma)))

import psycopg2

conn = psycopg2.connect(port=5432, user='chembl', dbname='chembl_17')

cur = conn.cursor()

sql1 = """
SELECT mr.*, md.chembl_id, cp.full_mwt, cp.alogp
from mols_rdkit mr, molecule_dictionary md, compound_properties cp
where
mr.m @> 'C1C[!#1!#6]1'::qmol
and
mr.molregno = md.molregno
and
md.molregno = cp.molregno
limit 100
"""

cur.execute(sql1)

for c in cur: print c

smi = 'CCCc1nn(C)c2C(=O)NC(=Nc12)c3cc(ccc3OCC)S(=O)(=O)N4CCN(C)CC4' #sildenafil

sql2 = """
select molregno,m as smiles,tanimoto_sml(morganbv_fp('CCCc1nn(C)c2C(=O)NC(=Nc12)c3cc(ccc3OCC)S(=O)(=O)N4CCN(C)CC4'::mol),mfp2) as similarity
from fps_rdkit join mols_rdkit using (molregno)
where morganbv_fp('CCCc1nn(C)c2C(=O)NC(=Nc12)c3cc(ccc3OCC)S(=O)(=O)N4CCN(C)CC4'::mol)%mfp2
order by morganbv_fp('CCCc1nn(C)c2C(=O)NC(=Nc12)c3cc(ccc3OCC)S(=O)(=O)N4CCN(C)CC4'::mol)<%>mfp2;
"""


cur.execute(sql2)

for c in cur: print c

from rdkit.Chem import PandasTools
import pandas as pd

data = pd.read_sql(sql2, conn)

data.shape

data.head()

nn = data[data['similarity'] >= 0.8]

nn.shape

nn.head()

PandasTools.AddMoleculeColumnToFrame(data,smilesCol='smiles',molCol='mol',includeFingerprints=True)

data.head()

data['logp'] = data['mol'].map(Descriptors.MolLogP)
data['mw'] = data['mol'].map(Descriptors.MolWt)

data.head()

data.sort(columns=['similarity']).head()

data[['mw', 'logp']].describe()

rcParams['figure.figsize'] = 12,12

data['logp'].hist()

scatter(data['mw'],data['logp'])

data[data['mw']>800][['molregno','mol','mw','logp']]

qsmi = 'c1nn(C)c2c1nc[nH]c2=O'

qmol = Chem.MolFromSmiles(qsmi)

qmol

subset = data[data['mol'] >= qmol]

subset.shape

subset.head()

data.groupby(data['mol'] >= qmol).describe().unstack()

data['containsQ'] = data['mol'] >= qmol

data.head(2)

data.boxplot('similarity',by='containsQ')

conn.close()