Optimization of multi-drug composition for the most efficacious action
Date
2010-06-02T19:53:07Z
Authors
Bose, Chris
Bizuet, Rocky
Caudillo, Luz
Gong, Jiafen
Jain, Rashi
Romanko, Oleksandr
Samarbakhsh, Abdi
Tam, Yun K.
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
In this report we consider a drug-design problem as it typically appears in Chinese medicine, where a large number of potentially active components are combined into one herbal therapy. The problem is posed by the SinoVeda Canada Inc. We focus on techniques for component-activity analysis in order to isolate the most active components, and the most active synergies between components as measured by tissue response to prepared herbal mixtures. The aim is to produce both qualitative and quantitative descriptions of the most active dose-fractions for a given herbal therapy and to aid in optimal design of herbal mixtures treating the target diseases. This report describes various techniques to achieve this goal, including multivariate linear regression, principle component analysis, subset selection and regularization.
Description
Keywords
Chinese medicine, drug-design, component-activity analysis, optimization, principal components analysis (PCA), multivariate linear regression, subset selection, least absolute shrinkage and selection operator (LASSO)