Hi All,
Recently I am working on a biomarker-discovering project through which relationship between change of human plasma's metabolites and drug treatment's toxicity is explored. We profiled the metabolites in the patient's plasma using metabolomics method and found more than 100 metabolites whose levels are significantly changed after drug treatment (compared to the pre-treatment levels). We also have toxicity data generated by clinical group recording all the side effects occurred in patients. I wanna suggestion for the best analytical methods for finding out which changes of metabolites are closely related to the occurrence of side effects. The one I plan to use is logistic regression (dependent variable is occurrence of a specific side effect, independent variables are the level changes of the different metabolites). I am not sure whether this method is right one or not.
In addition, is there any way to cluster the two different types of data (side effects and level change of metabolite) and generate the heatmap in one figure with the above map is for toxicity and the below map is for level change of metabolite. The x-axis for the both maps is patient's code number.
Please forward me the links if you know any publication related to the research that is similar to mine. I desperately need to know the analytic methods and the way to present the data.
Your prompt response will be highly appreciated.
Jeff
