Resources

Algorithms

Fingerprinting is a common technique for molecular screening. Daylight, Inc. introduced this technique, and it is described in the following article. Bingo fingerprints, as compared to Daylight fingerprints, are built not from bond paths, but from trees and rings. A Russian article describes the enumeration of subtrees, which in turn is based on reverse search by David Avis and Komei Fukuda.

For tautomer substructure search, Bingo incorporates a unique type of fingerprinting.

For subgraph matching, we developed an original algorithm. It is somewhat similar to an algorithm by Luigi Cordella et al.

For the molecule layout in Indigo and Bingo, we developed a unique algorithm, based on biconnected components extraction. It incorporates some ideas of Craig Shelley.

For the aromaticity matching and resonance search, we developed unique algorithms as well.

For the canonical SMILES, Brendan McKay's nauty canonical labeling algorithm was re-implemented. Many of the original nauty's features and optimizations were not included. There is also a description of the algorithm in Russian.

For the affine transformation matching, Wolfgang Kabsch's algorithm was implemented.

In internal molecule and reaction formats, the LZW algorithm with some modifications was used for compression.

For the exact maximum common substructure search, we developed a unique algorithm partially based on Thierry Hanser's algorithm (which in turn has its roots in an algorithm by Coen Bron and Joep Kerbosh). For the approximate maximum common substructure search, the 2DOM algorithm with some modifications was implemented. A Russian paper describes both algorithms.

Oracle Interface

The C++ wrapper was implemented for the Oracle OCI library. One Russian article describes the pitfalls of handling Oracle LOBs in OCI, while another Russian article covers the performance of queries.

Cairo

Cairo 1.8.6 was used for rendering in Indigo. In Linux builds, the system-wide installation of libcairo must be present (-I/usr/include/cairo flag is passed to the compiler and -lcairo is passed to the linker). In Mac OS X builds, the system-wide installation of libcairo must be present as well (/opt/local/lib/libcairo.dylib path is passed to the linker). You can install it using MacPorts. The 32-bit and 64-bit Windows builds of Cairo are included in the source tree. They can be downloaded from GTK and GNOME project sites, respectively. Binaries of libpng, zlib, Freetype, Fontconfig, and Expat, needed for Cairo, are there as well.

File Formats

The molecule and reaction format descriptions are available on the following sites: Daylight SMILES, ChemAxon SMILES extensions, and MDL (Symyx) Molfiles.

Web standards

More information about web standards used by Ketcher can be found here: SVG, VML.

Web frameworks

Ketcher includes two JavaScript libraries: Raphaël for vector graphics and Prototype for overall code improvement.

Code Examples

Chemistry Toolkit Rosetta wiki contains some examples of small utilities written using Indigo C++ API.

Development Tools

We use NetBeans 6.8 on Linux workstations, Microsoft Visual Studio 2008 on Windows workstations, and XCode 3.2 on Mac OS X workstations.

Commercial availability

We have dual-licensed our code. If the GPL-licensed code does not fit your needs, please contact us at info@ggasoftware.com to discuss the purchase of a commercial license. You may need the commercial license if you want to:

  • Receive ongoing support and maintenance of our code
  • Include our code in your proprietary software product