The thoughts, ideas, help, and support of many people went into making
this project. In no particular order they are:

Noel O'Boyle: I used his cinfony code as the base to understand how
the different projects generate fingerprints.  He also encouraged me
to develop these tools and put them out to the world.

Geoff Hutchison for pointing out that he most often works with small
data sets, and human-readable/text formats are more important than a
binary format which is designed for million-compound data
sets. (Thinking about this was the source for developing both a text
and a binary format.)

The OpenBabel developers, for developing a widely used tool in
cheminformatics with both hash and substructure fingerprints.

Greg Landrum, for RDKit and for discussions about what's needed in a
fingerprint format. I hope some day to include some features he needs,
like sparse and count fingerprints. RDKit's MACCS definitions are
(excepting minor portability tweaks) the definitions used for the
RDMACCS patterns. Greg's definitions are also used in CDK and
OpenBabel.

OpenEye, for a non-commercial license to their tools.

Phil Evans, for being the first external person to test the code and
give feedback about installation and Python 2.5 compatibility
problems.

Evan Bolton and Wolf-Dietrich Ihlenfeldt for their help in
understanding the PubChem/CACTVS substructure keys.

Rajarshi Guha for implementing the PubChem/CACTVS substructure keys in
CDK (and the anonymous person at the NIH Chemical Genomics Center who
developed the precursor version). The CDK's independent implementation
helped me cross-check my implementation. He also added support for the
FPS format in his fingerprint analysis tools.

Dmitry Pavlov for insights in how Indigo works and improvements in the
Indigo implementations.

To patches and feedback from Chris Morely, Jörg Kurt Wegner, Phil
Evans, Björn Grüning, Andrew Henry and others.

To you, for reading this list and using the code.
