The DNA microarray technology has contributed widely in biological studies. Recently, a method to detect the effects of extracellular matrix (ECM) proteins on cells using microarray was developed. In this dissertation, the microarray technology was applied to study the effects of ECM proteins on human mesenchymal stem cell (hMSC) differentiation, and to identify the best growth- supporting ECM conditions for glioblastoma, and to create a method for efficient small molecule drug screening. The importance of ECM on hMSCs is well known. Numerous studies have shown that the ECM deposed by specific types of cells could induce differentiation in hMSCs. However, clinical application of cell-deposited ECM could carry risks such as contamination, immune rejection and batch-to-batch variation. An artificial ECM could avoid these problems. In this dissertation, the effects of ECM proteins were screened on the hMSCs differentiation using the microarray technology. The myogenic and neurogenic differentiation profiles are reported. This information can contribute to the selection of optimal differentiation-inducing coating on the muscle or nerve tissue engineering scaffolds. Cell culture has always been a fundamental tool in cancer research. However, most primary cancer cells are very difficult to maintain. In this dissertation, ECM protein combinations were tested for their efficacy in supporting the growth of glioblastoma cells. Twelve best combinations were identified. The result may lead to successful maintenance of the primary cells of glioblastoma and other types of cancerous tumors. The results may yield larger quantities of the primary cancer cells, which are heretofore limited in supply, and thus facilitate the advancement of cancer research. Genetic marker-based personalized medicine has attracted much attention. However, our understanding of the genes is still limited. A direct drug efficacy screening on the diseased cells of an individual patient may complement the gene-based personalized medicine. In this study, a small-molecule drug microarray method is reported. This method is able to screen a large number of drugs and their combinations by using a very small mount of patients' diseased cells, making it possible to design personalized drug cocktails