A Bayesian binary regression analysis was used to generate a predictor of oxaliplatin sensitivity from the gene expression data.
With the Gaussian vector space kernel, we can apply a kernel-based method to generate a predictor of any one of several biological activities, for example: ADMET properties or affinity of ligands used as therapeutic agents.
In this article, we integrated and leveraged information from different sources, such as chemical similarity, targets and adverse drug effects (ADEs), to generate a predictor to identify drug-targets, target-adverse effects, and drug-adverse effects associations.
To determine if the gene signatures gave prognostic information additional to the clinical variables described above, we generated a predictor based on a principal components analysis using the clinical variables, and then developed additional models using both the clinical variables and the individual gene signatures.
For each patient we generate a predictor vector whose k components consist of the differences of signal intensities from tissue samples from k gene pairs.
A distinct subset of these generate a predictor of future cardiovascular death that replicates in two phases of analysis and is strongly associated with survival time in the combined dataset.
As a validation of our overall approach, we also tested the utility of generating a breast tumor predictor from the human PBMC gene expression data using the same parameters applied to the mouse gene expression data.
Both modules were linked to generate a final predictor that establishes hypothesis about new drug-target-adverse effect candidates.
If independently validated, a peripheral blood transcript abundance test has potential, in conjunction with measures of cholesterol, serum creatinine, and white blood cell counts, to generate a significant predictor of risk of cardiovascular death in patients with CAD.
Genes within 20 kb of binding sites belonging to cluster C1 and whose expression was impacted by FOXM1 knockdown were used to generate a gene predictor that was employed to interrogate two large, independent data sets [ 38, 41] of ERα-positive breast cancer patients treated with TAM.
