This project builds the program smg (SMARTS Multiplet Generator). It generates what might loosely be called 2D pharmacophore fingerprints. In this case, it's a file of 1s and 0s for each input molecule, denoting the presence or absence of a particular pharmacophore grouping. Smg can output sites, pairs or triplets. Sites are the individual pharmacophore features, pairs are combinations of 2 features and a through-bond distance (i.e. the counts of the number of bonds between the two features via the shortest path) and triplets are three features with through-bond distances.
An example pair would be acceptor:3:posch i.e. an h-bond acceptor 3 bonds from a positive charge.
An example triplet would be acceptor:10:acceptor-acceptor:11:rings-acceptor:12:rings which describes 2 h-bond acceptors and a ring 10, 11 and 12 bonds apart. When calculating the distance of a ring to another feature, it is always taken as the nearest ring atom to the feature, so the distance is always the shortest possible.
The pharmacophore features are described using a set of SMARTS definitions and a description of how the SMARTS patterns should be combined to form the feature. An example of each input file is in the test_dir directory, as test.smt and test.points respectively.
In the output file, the columns are headed by a short, unintelligble label. This a hashed version of the full name for the feature. A subsidiary file with suffix name_decode is made which provides a translation. This was done to keep the file of a manageable size, and dates from the days when people at AZ were analysing stuff using SAS which had a relatively short upper limit on the length of a column name.
Requires: a recent version of OEChem.
To build it, use the CMakeLists.txt file in the src directory. It requires the following environment variable to point to a relevant place:
OE_DIR - the top level of an OEChem distribution
Then cd to src and do something like: mkdir dev-build cd dev-build cmake -DCMAKE_BUILD_TYPE=DEBUG .. make
If all goes to plan, this will make a directory src/../exe_DEBUG with the executables in it. These will have debugging information in them.
For a release version: mkdir prod-build cd prod-build cmake -DCMAKE_BUILD_TYPE=RELEASE .. make
and you'll get stuff in src/../exe_RELEASE which should have full compiler optimisation applied.
These instructions have only been tested in Centos 6 and Ubuntu 14.04 Linux systems. I have no experience of using them on Windows or OSX, and no means of doing so.
David Cosgrove AstraZeneca 12th February 2016
davidacosgroveaz@gmail.com