Abstract: An organized knowledge-supervised approach—Multidimensional Integrative eXpression Profiling (MIXP)—can not only improve sample classification accuracy by serving as a feature transformation approach, but also help in the discovery of groups of crucial molecular entities that have been too weak to detect individually through preexisting methods. Functionally related molecules that are individually expressed with low differentials, have often been considered as noise and ignored in traditional studies, but through the MIXP approach, they can be readily identified by virtue of their coordinate expression.

Abstract: An integrative pathway modeling approach and ranking/evaluating algorithms based on disease-specific pathway models can predict drug efficacy for patients based on their gene expression profiles. A disease-specific pathway model is first constructed with proteins and drugs important to the disease by using computational connectivity maps (C-Maps). Through the pathway model-based ranking algorithm, ideal drugs or optimized drug combination can be discovered for a patient to modulate the gene expression profile of this patient close to those in healthy individuals at pathway-level.

Abstract: A computational systems pharmacology framework consisting of statistical modeling and machine learning based on comprehensive integration of systems biology data, including drug target data, protein-protein interaction (PPI) networks, and gene ontology (GO) annotations, and reported drug side effects, can predict drug toxicity or drug adverse reactions (ADRs). Biomolecular network and gene annotation information can significantly improve the predictive accuracy of ADR of drugs under development. The use of PPI networks can increase prediction specificity, and the use of GO annotations can increase prediction sensitivity.