Materials genomics approaches emphasize data production, data organization, and new types of data analysis based on large databases of information. In this talk, I will present several new open-source codebases being developed at LBNL to aid researchers in all aspects of this process. The AMSET code and method models various electronic scattering processes; its application to thermoelectrics materials discovery will be discussed. The FireWorks and atomate codebases allow one to automatically generate workflows that calculate multiple materials properties, scale those computations to millions of jobs at supercomputing clusters, and parse results into searchable databases. Next, I will introduce matminer, which is a code that is specifically aimed at data extraction, data analysis, and machine learning of crystalline materials properties. Finally, I will describe our work in auto-optimization of materials parameters in situations where forward simulations are possible but expensive and the inverse solution is desired.