Meta-QSAR: a large-scale application of meta-learning to drug design and discovery
- Submitting institution
-
Brunel University London
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 048-181811-15394
- Type
- D - Journal article
- DOI
-
10.1007/s10994-017-5685-x
- Title of journal
- Machine Learning
- Article number
- -
- First page
- 285
- Volume
- 10
- Issue
- 7
- ISSN
- 0885-6125
- Open access status
- Compliant
- Month of publication
- December
- Year of publication
- 2017
- URL
-
https://link.springer.com/article/10.1007/s10994-017-5685-x
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
6
- Research group(s)
-
1 - Artificial Intelligence (AI)
- Citation count
- 19
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Investigates the learning of quantitative structure activity relationships (QSARs) as a case-study of meta-learning. This area is of high societal importance, and a key step in the development of new medicines. Although many machine learning methods have been applied to QSAR learning there is no agreed best way, and therefore the problem is well-suited to meta-learning. We carried out the most comprehensive ever comparison of machine learning methods: 18 regression methods, 3 molecular representations, applied to more than 2,700 QSAR problems. The curated QSAR dataset has been further used in a study by MIT and Stanford published in ICML2019 http://proceedings.mlr.press/v97/diakonikolas19a.html.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
