Discerning and exploiting patterns in chemical data lies at the heart of any systematic program for materials design. Rapid advances in computational power during the last couple of decades have enabled theoretical characterization of many bio- and nano-materials using first principles computations, but the computational effort required is still formidable enough to preclude routine use of such methods in a high-throughput setting. Many problems in chemistry and materials science are so complex that novel approaches based on the sciences of informatics, coupled with first principles computations, are required for their solution. Predictive informatics methods employing statistical techniques to process and analyze the huge volumes of data generated by robotic high-throughput assays in the wake of the Human Genome Project led to rapid advances in bioinformatics. In the search for materials with specific properties, a similar combination of first principles computational studies with experimental work and heuristic statistical methods has the potential to leverage the power of each, thereby bringing high-throughput capability to the quest for predicting the characteristics of materials at the nanoscale, with the aim of accelerating the development of new materials with specific properties. This approach has been termed the Materials Genome Initiative. The chemistry of clusters of small numbers of atoms and of nano alloys are test bed applications where a materials genome approach shows promise of yielding a quantitative breakthrough in our understanding of these materials and quickly leading to interesting applications.