1. Scientific & technical presentation JChem Cartridge for Oracle
version 5.3, January 2010 1 1 1

2. Contents Purpose of JChem Cartridge Features of JChem Cartridge
Constituents of the JChem Cartridge API
Normal Tables vs. JChem Tables
Architecture of JChem Cartridge 2 2 2

3. Purpose of JChem Cartridge
Access JChem functionality using SQL:
SELECT count(*) FROM nci WHERE jc_contains(structure, 'Brc1cnc2ccccc12') = 1
Access JChem in any programming environment offering Oracle connectivity (.NET, Java, Perl, PHP, Python, Apache mod_plsql...)‏
Execute SQL queries efficiently using extensible indexes
Precompute chemical information on structures by creating jc_idxtype indexes:
CREATE INDEX jcxnci ON nci(structure) INDEXTYPE IS jc_idxtype
The jc_idxtype implementation scans the indexed column for eligible structures in one single performance-optimized operation: domain index scan 3 3 3

4. Features of JChem Cartridge
Adds chemistry knowledge into the SQL language of Oracle (SELECT, INSERT, UPDATE, ...)‏
Substructure, superstructure, full structure, similarity searching
Complex chemical expressions using the Chemical Terms language that includes logP, pKa, ...
Automatic property calculation during registration
Standardization (canonicalization) during registration
Structure format conversions (MRV, Molfile, SDfile, RDfile, SMILES, CML, etc.)‏; 2D, 3D image generation
Structure enumeration using reaction rules
User-defined fingerprint columns
Custom similarity search through molecular descriptors
Interaction with Oracle optimizer 4 4 4

5. Structure search features
Wide range of query atoms
Query properties
R-group queries
Full SMARTS support
Coordination compounds
Link nodes
Pseudo atoms, lone pairs
Relative stereo
Reaction search features
Hit coloring, position variation
Polymers
See detailed information on structure search: 5

6. Search options Stereo on/off
Ignore charge/isotope/radical/valence/mixture brackets
Vague bond matching options
Chemical Terms filter
Tautomer search
Inverse hit list
Maximum search time / number of hits
Combine with non-structure conditions
Ordering of results
etc. 6

7. Searching in Markush structures
Combinatorial Markush structure registration and search
Markush features handled in search & enumeration:
R-groups (nesting to any depth)
Atom lists, bond lists
Position variation bond
Link nodes and repeating units
Homology groups
Compatible Markush enumeration plugin
Detailed description: 7 7

8. Standardization Default standardization includes: Hydrogen removal
Aromatization
Custom standardization can be specified for each table or JChem index 8

9. Custom Standardization Example
JChem Cartridge
Custom Standardization Example
before
after 9

10. Compatibility and integration
File formats:
SMILES
MDL molfile (v2000 and v3000)‏
MDL SDF
RXN
RDF
MRV
IUPAC name, InChI
Markush DARC
CDX
Operating systems:
Windows
Linux
Solaris
HP-UX
etc.
DB engines:
Oracle versions 9i R2 or above
for alternative RDBMS systems, see the JChem Base presentation: 10

11. Elements of the JChem Cartridge API
Operators (jc_...) for SQL and their functional forms (jcf package) for PL/SQL
Parameters for index creation
DML operators for JChem tables
Support functions for user defined operators 11 11 11

12. Operators and functions I.
Typical operator:
jc_<some-operation>(<target-structure-column>, <some- operand>)‏
Operator for substructure search:
jc_contains(<target-structure-column>, <query-structure>)‏
“Swiss-army-knife” search operator:
jc_compare(<target-structure-column>, <query-structure>, <options>)‏ 12 12 12

13. Operators and functions II.
Chemical Terms
Over 100 built-in functions, including
elemental analysis
topological descriptors
property predictions (logP/D, pKa, PSA, H bond donors/acceptors, charge etc).
tautomers, protonation forms
User-defined functions.
Example: The Lipinski-rule in chemical terms
SELECT count(*) FROM nci_3m WHERE jc_compare(structure, 'O=C1ONC(N1c2ccccc2)-c3ccccc3','sep=! t:s!ctFilter:(mass() <= 500) && (logP() <= 5) && (donorCount() <= 5) && (acceptorCount() <= 10)') = 1 13 13 13

14. Operators and functions III.
jc_compare: substructure/similarity/exact searching combined with Chemical Terms expressions
jc_matchcount: number of occurences of the query structure in the target
jc_evaluate: Chemical Terms evaluation
jc_molweight: molecular weight
jc_formula: molecular formula
jc_react: structure enumeration based on virtual reactions
jc_standardize: structure canonicalization
jc_molconvert: conversion to different formats (image generation is supported)
jc_tanimoto: similarity search
jcf.hitColorAndAlign: substructure coloring and alignment 14 14 14

15. Operators and functions IV.
Similarity search example displaying ID, SMILES code, and molweight:
SELECT cd_id, cd_smiles, cd_molweight FROM my_structures WHERE jc_tanimoto(cd_smiles, 'CC(=O)Oc1ccccc1C(O)=O') >= 0.8;
Chemical Terms and Query Prefiltering:
SELECT id, purchase_date FROM compounds_instock WHERE jc_compare(structure, 'C(=S)([N][N])[S]', 'sep=! t:t!simThreshold:0.9!ctFilter:logp()>1!filterQuery:SELECT rowid FROM compounds_instock WHERE purchase_date > DATE '' ''') = 1
Prefiltering allows to execute search on a subset of rows more efficiently.
Dynamic generation of static images:
SELECT jc_molconvertb(structure, 'png -2') FROM nci where id = :1
Avaliable image formats: png, jpeg, svg, ...
PNG 15 15 15

16. Operators and functions V.
Calculate logp:
SELECT jc_evaluate('OC(=O)c1c2ccccc2nc3ccccc13', 'logp') FROM dual;
Generate tautomers:
SELECT jc_evaluate_x('NC1=C(CC=O)C=CCC1', 'chemTerms:tautomers() outFormat:smiles') FROM dual;
Generate resonants:
SELECT jc_evaluate_x('NC1=C(CC=O)C=CCC1', 'chemTerms:resonants() outFormat:smiles') FROM dual 16 16 16

17. Index parameters Index parameters affect: Examples:
Fingerprint attributes
Standardizer configuration
Table space and storage options of the index table
Examples:
Standardization by stripping hydrogens and using basic aromatization:
CREATE INDEX jcxnci ON nci(structure) INDEXTYPE IS jc_idxtype PARAMETERS('STD_CONFIG=removeexplicitH..aromatize:b')‏
Add structural keys to fingerprint for more efficient substructure searching (structural keys are defined in table stfp_keys):
CREATE INDEX jcxnci ON nci(structure) INDEXTYPE IS jc_idxtype PARAMETERS('STRUCTURALFP_CONFIG=select structure from stfp_keys')‏ 17 17 17

18. Calls Not Using Indexes
Using SQL statements for calling JChem operators on structures not stored in a table
Sample SQL statement without index information:
SELECT jc_contains('O=C1C=CNC=C1', 'n1ccccc1') FROM dual
Setting default properties for calls not using indexes:
CALL jc_set_default_property('standardizerConfig', 'aromatize:b')‏ 18 18 18

19. Supported Column Types
VARCHAR2: typically for short formats, e.g. SMILES
CLOB
BLOB
for longer formats, e.g. MDL molfile,
Marvin (mrv)‏ 19 19 19

20. Supported Structure Table Types
Regular Table: nci_1k
JChem Table (generated by jcman or API): jc_nci_1k
CREATE INDEX jcxnci_1k...
Rowid of the base table (nci_1k)‏
Index table:
jcxnci_1k_jcx
CREATE INDEX jcxjc_nci_1k... 20 20 20

21. Regular Tables vs. JChem Tables
Regular structure tables
base table and index table are physically distinct
index properties are specified as index parameters
JChem structure tables
base table and index table are physically the same
most of the “index” properties are specified during table creation (jcman or Java API)‏
Pros & Cons:
inserts from outside the database are faster with JChem tables
JChem tables require Java API or the jcman command line tool (for table creation) and Java API or special cartridge functions for INSERTs, UPDATEs and DELETEs; standard SQL can be used with regular tables in all cases. 21 21 21

22. JChem Cartridge Architecture
Computation intensive operations are performed in a separate Sun JVM.
Advantage: fast execution (optimized native code)‏
flexibility in deployment
JChem Server
Oracle
JChem Cartridge
JChem Streams
RMI
JChem Base
Search
Update
Cache
Cache
JDBC
JChem Core 22 22 22

23. Performance
Table containing 19,528,372 structures from PubChem with Intel Quad CPU Q GHz desktop PC, 8GB memory desktop PC
Substructure search results:
Query Structure
Hit Count
Time (ms)‏
C1CN1c2cnnc3c(cncc23)C4=CSC=C4
1487
O=C1ONC(N1c2ccccc2)c3ccccc3
129
823
Oc1c(N=N)c(cc2cc(ccc12)S(O)(=O)=O)S(O)(=O)=O 93
764
C(Sc1ncnc2ncnc12)c3ccccc3
489
786
NC1=CC=NC2=C1C=CC(Cl)=C2
6,001
1,189
c1ncc2ncnc2n1
146,256
6,665
Clc1ccccc1
2,975,285
82,646
JChem 5.2 23 23 23

24. Future plans Flexible 3D pharmacophore search R-Group decomposition
Clustering
Maximum common substructure search type
Extended fingerprint connectivity (EFPC) 24 24 24

25. Summary
JChem Cartridge for Oracle allows to access the rich functionality of JChem Base in a flexible and efficient manner.
JChem Cartridge for Oracle uses creative solutions to broaden the applicability of JChem's core functions while preserving key benefits of the Java platform. 25 25 25

26. Links Documentation Forum Brochure
Forum
Brochure 26 26 26

27. Visit other technical presentations
MarvinSketch/View
MarvinSpace
Calculator Plugins
JChem Base
JChem Cartridge
Standardizer
Screen
JKlustor
Fragmenter
Reactor 27 27 27