The model plant Arabidopsis thaliana encodes 13 putative oxidosqualene cyclases (OSC), and the product profiles of some of these have been determined. Described herein are experiments to characterize the products of the oxidosqualene cyclase PEN1 (encoded by At4g15340) from Arabidopsis. PEN1 was PCR-amplified from an A. thaliana young seedling cDNA library and heterologously expressed in a yeast strain that was metabolically engineered to accumulate the substrate oxidosqualene. This recombinant system accumulated cyclized materials, which were characterized by GC-MS, GC and NMR analysis. PEN1 is the most prolific known oxidosqualene cyclase, making more than 30 distinct triterpene alcohols, or more than a third of the triterpene skeletons that have been isolated from plants. The predominant compound is a new tricyclic triterpene diol, the structure of which was established using ¹H and ¹³C 1D NMR, COSY HSQC, HMBC, NOE and NOESY 2D experiments. A challenging component of the triterpene diol structure determination was solved by chemical modification of the side chain. The genome mining approach is generating a total triterpene alcohol profile of the plant, and has led to the discovery of large variety of novel triterpenes. A correlation of sequence and product is facilitating further elucidation of the relationship between protein structure and product triterpenes.