- Matt Robinson, [email protected]
This benchmarking platform was originally created by Riniker and Landrum and described in:
S. Riniker, G. Landrum, J. Cheminf., 5, 26 (2013), DOI: 10.1186/1758-2946-5-26, URL: http://www.jcheminf.com/content/5/1/26
The dataset consists of 88 targets from MUV, DUD, and ChEMBL.
The original code can be found at https://github.com/rdkit/benchmarking_platform/blob/master/analysis/analysis_functions.py
The original platform was written in python 2, and was found to be somehwhat difficult to run. Consequently, we have updated the code to be both python 2 and 3 compatible, for ease of use. Futherrmore, we have changed the structure of the code to facilitate benchmarking without the need for navigating the complicated file structure.
The original code is contained in the subdirectory benchmarking_platform_p3. Most of the code is unchanged excepting changes to allow for python 3 usage. However, several files have been changed more significantly in order to simplify usage. For example, most scripts have been changed to accept abolute -- instead of relative paths -- as parameters.
The easiest way to get started is to clone this GitHub repository. Please then go to the data/README.md file for instructions on installing the data.
All code should be run in the provided conda environment
This environment can be installed using
conda env create -f environment.yml
To modify the specific parameters of the benchmark, please visit the BENCHMARK_CONFIG.md file and change the desired parameters.
After this configuration file is appropriately modified, run the benchmark using the run_benchmarks.py script. Finally, an automatically generated markdown report of the results can be produced using produce report.py.
An overview of these steps is below:
- Follow insructions to download data found in
data/README.md - Make sure the cas-env conda environment is active
- Modify
BENCHMARK_CONFIG.mdas desired python run_benchmarks.pypython produce_report.py
The generated plots are found in benchmark_results_plots and the report is located at benchmark_report.md.
Please see and modify benchmarking_platform_p3/scoring/fingerprint_lib.py. The file contains the necessary instructions.
The original benchmarking code was distributed with the following license
Copyright (c) 2013, Novartis Institutes for BioMedical Research Inc. All rights reserved.
Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following
disclaimer in the documentation and/or other materials provided
with the distribution.
* Neither the name of Novartis Institutes for BioMedical Research Inc.
nor the names of its contributors may be used to endorse or promote
products derived from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
This licesnse does not cover the CAS fingprints, which are a product of CAS.