aschreyer

    WITH conformer AS
         (
          SELECT openeye.omega('CC1=CC(=C(N1C2=CC(=CC(=C2)C(F)(F)F)C(F)(F)F)C)/C=C\3/C(=O)N=C(S3)NC4=CC=CC=C4') as oeb
         ),
         query AS
         (
          SELECT openeye.usrcat(oeb) AS usr_moments
            FROM conformer
         ),
         screen AS

aschreyer

import requests
 
headers = {'accept': 'application/json'}
 
files = {'file': open('/home/adrian/Dropbox/OSM-S-35-I.sdf')}
 
response = requests.post('http://marid.bioc.cam.ac.uk/credo/chemcomps/sim/usr/', 
                         headers=headers, files=files, params={'usrcat':'t', 'limit':25})

aschreyer

loop_
_pdbx_struct_oper_list.id 
_pdbx_struct_oper_list.type 
_pdbx_struct_oper_list.name 
_pdbx_struct_oper_list.symmetry_operation 
_pdbx_struct_oper_list.matrix[1][1] 
_pdbx_struct_oper_list.matrix[1][2] 
_pdbx_struct_oper_list.matrix[1][3] 
_pdbx_struct_oper_list.vector[1] 
_pdbx_struct_oper_list.matrix[2][1] 

aschreyer

-- set search_path = public, openeye;
 
DO $$
    DECLARE
        molregno integer;
        fp oefp;
    BEGIN
        FOR molregno, fp IN   SELECT cfp.molregno, cfp.circular_fp
                                FROM chembl.compound_oefps cfp
                            ORDER BY random() LIMIT 100

aschreyer

@paginate
def fetch_all_descendants(self, fragment_id, *expr, **kwargs):
    """
    Returns all the descending fragments of the fragment with the given
    fragment_id using an recursive SQL query.
    """
    # query part that will be used in both parts of the recursive query
    query = Fragment.query.with_entities(Fragment.fragment_id)
    query = query.join(FragmentHierarchy, FragmentHierarchy.child_id==Fragment.fragment_id)

aschreyer

@hybrid_method
@requires.rdkit
def contains(self, smiles):
    return self.rdmol.HasSubstructMatch(MolFromSmiles(str(smiles)))

@contains.expression
def contains(self, smiles):
    return self.rdmol.op('OPERATOR(rdkit.@>)')(smiles)

aschreyer

import urllib2
import json

headers = {'Accept' : 'application/json'}
req  = urllib2.Request('http://marid.bioc.cam.ac.uk/credo/chemcomps/001', headers=headers)
resp = urllib2.urlopen(req)
jsondata = json.load(resp)

# output
{u'data': [7464,

aschreyer

  SELECT assay_id, 
         count(DISTINCT molregno) AS compounds, 
         openeye.multimcs(array_agg(ism), threshold:=0.8) AS mcs
    FROM chembl.target_dictionary
    JOIN chembl.assay2target USING(tid)
    JOIN chembl.activities a USING(assay_id)
    JOIN chembl.compound_smiles cs USING(molregno)
   WHERE protein_accession = 'P53779' 
GROUP BY assay_id
  HAVING count(DISTINCT molregno) >= 5

aschreyer

SELECT * 
  FROM credo.unichem_pdb_to_kegg
 LIMIT 10;

het_id compound_id 
------ -----------
LQZ    C07073
HC4    C00811
DC5    C14355
POD    C10874

aschreyer

-- select a query molecule
WITH query AS
(
    SELECT * 
      FROM dude.usr_moments_fnta 
     WHERE category = 'ACTIVE' AND mol_idx = 1 AND conformer = 0
)
-- run the virtual screen with the given query
  SELECT category, mol_idx, MAX(sq.similarity) as similarity
    FROM (