A structure prediction tool has been developed to guide the discovery of MOF materials. This computational strategy has been trained over a series of existing MOFs and further successfully applied in tandem with an experimental effort to produce novel Zr MOFs based on naturally occurring carboxylic acids.