Bayesian optimization for conformer generation
Abstract Generating low-energy molecular conformers is a key task for many areas of computational chemistry, molecular modeling and cheminformatics. Most current conformer generation methods primarily focus on generating geometrically diverse conformers rather than finding the most probable or energ...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2019-05-01
|
| Series: | Journal of Cheminformatics |
| Subjects: | |
| Online Access: | http://link.springer.com/article/10.1186/s13321-019-0354-7 |
| _version_ | 1811331852024152064 |
|---|---|
| author | Lucian Chan Geoffrey R. Hutchison Garrett M. Morris |
| author_facet | Lucian Chan Geoffrey R. Hutchison Garrett M. Morris |
| author_sort | Lucian Chan |
| collection | DOAJ |
| description | Abstract Generating low-energy molecular conformers is a key task for many areas of computational chemistry, molecular modeling and cheminformatics. Most current conformer generation methods primarily focus on generating geometrically diverse conformers rather than finding the most probable or energetically lowest minima. Here, we present a new stochastic search method called the Bayesian optimization algorithm (BOA) for finding the lowest energy conformation of a given molecule. We compare BOA with uniform random search, and systematic search as implemented in Confab, to determine which method finds the lowest energy. Energetic difference, root-mean-square deviation, and torsion fingerprint deviation are used to quantify the performance of the conformer search algorithms. In general, we find BOA requires far fewer evaluations than systematic or uniform random search to find low-energy minima. For molecules with four or more rotatable bonds, Confab typically evaluates $$10^{4}$$ 104 (median) conformers in its search, while BOA only requires $$10^{2}$$ 102 energy evaluations to find top candidates. Despite using evaluating fewer conformers, 20–40% of the time BOA finds lower-energy conformations than a systematic Confab search for molecules with four or more rotatable bonds. |
| first_indexed | 2024-04-13T16:27:04Z |
| format | Article |
| id | doaj.art-bdac2ad431034711afcb83fea6fe8d7d |
| institution | Directory Open Access Journal |
| issn | 1758-2946 |
| language | English |
| last_indexed | 2024-04-13T16:27:04Z |
| publishDate | 2019-05-01 |
| publisher | BMC |
| record_format | Article |
| series | Journal of Cheminformatics |
| spelling | doaj.art-bdac2ad431034711afcb83fea6fe8d7d2022-12-22T02:39:43ZengBMCJournal of Cheminformatics1758-29462019-05-0111111110.1186/s13321-019-0354-7Bayesian optimization for conformer generationLucian Chan0Geoffrey R. Hutchison1Garrett M. Morris2Department of Statistics, University of OxfordDepartment of Chemistry and Chemical Engineering, University of PittsburghDepartment of Statistics, University of OxfordAbstract Generating low-energy molecular conformers is a key task for many areas of computational chemistry, molecular modeling and cheminformatics. Most current conformer generation methods primarily focus on generating geometrically diverse conformers rather than finding the most probable or energetically lowest minima. Here, we present a new stochastic search method called the Bayesian optimization algorithm (BOA) for finding the lowest energy conformation of a given molecule. We compare BOA with uniform random search, and systematic search as implemented in Confab, to determine which method finds the lowest energy. Energetic difference, root-mean-square deviation, and torsion fingerprint deviation are used to quantify the performance of the conformer search algorithms. In general, we find BOA requires far fewer evaluations than systematic or uniform random search to find low-energy minima. For molecules with four or more rotatable bonds, Confab typically evaluates $$10^{4}$$ 104 (median) conformers in its search, while BOA only requires $$10^{2}$$ 102 energy evaluations to find top candidates. Despite using evaluating fewer conformers, 20–40% of the time BOA finds lower-energy conformations than a systematic Confab search for molecules with four or more rotatable bonds.http://link.springer.com/article/10.1186/s13321-019-0354-7Bayesian optimizationGaussian processesConformer generationRotatable bondTorsion angleConformational space |
| spellingShingle | Lucian Chan Geoffrey R. Hutchison Garrett M. Morris Bayesian optimization for conformer generation Journal of Cheminformatics Bayesian optimization Gaussian processes Conformer generation Rotatable bond Torsion angle Conformational space |
| title | Bayesian optimization for conformer generation |
| title_full | Bayesian optimization for conformer generation |
| title_fullStr | Bayesian optimization for conformer generation |
| title_full_unstemmed | Bayesian optimization for conformer generation |
| title_short | Bayesian optimization for conformer generation |
| title_sort | bayesian optimization for conformer generation |
| topic | Bayesian optimization Gaussian processes Conformer generation Rotatable bond Torsion angle Conformational space |
| url | http://link.springer.com/article/10.1186/s13321-019-0354-7 |
| work_keys_str_mv | AT lucianchan bayesianoptimizationforconformergeneration AT geoffreyrhutchison bayesianoptimizationforconformergeneration AT garrettmmorris bayesianoptimizationforconformergeneration |