Methods, systems and computer readable media for discovering a combination of treatments to reduce the progress of, or eliminate a tissue malady. Gene expression values of at least one sample of tissue exhibiting the tissue malady and at least one reference sample tissue that does not exhibit the malady are measured using at least one CGH array designed to measure gene sequences and possible variations in gene sequences attributable to the malady. Gene expression signatures are generated from differential expression values of ratios of the measured gene expression values between the at least one sample exhibiting the malady and the at least one reference sample, across all samples, respectively. The tissue samples exhibiting the malady are treated with a treatment, and a treatment-response value is measured with respect to each of the tissue samples treated, as effected by the treatment. A phenotypic signature representing the treatment-response values of each of the tissue samples treated is generated for characterizing the effects of the treatment on the tissues treated. Processing may be repeated with a different treatment at least once so that multiple phenotypic signatures have been generated for multiple treatments. A clustering operation is then based on the gene expression signatures of the differential expression levels and the phenotypic signatures of the treatment-response values together, and treatments are selected by identifying the treatment-response phenotypic signatures caused by those treatments, and which are clustered with gene expression signatures representing differential expression levels representative of the at least one tissue sample exhibiting the malady.