Juan Cuerva and co-workers from the University of Granada in Spain have reported in JACS on a titanocene mediated polyolefin cyclization reaction that studied with Gaussian calculations. The basis set used to approximate the titanocene catalyst was 6-31G*. From my little knowledge of Gaussian, when using QM calculations to study a reaction mechanism, you need two things: 1) the basis set and 2) the method. The basis set is the linear combination of a set of functions used to approximate the molecular orbitals. There are many kinds of basis sets that can be used for Gaussian. In my experience with transition metals however, I was unable to use the 6-31G* basis set because the transition metals have more complicated molecular orbitals since they have d-orbitals. I had to use the Los Alamos basis set: LANL2Dz. I eventually gave up on my time with Gaussian because I just kept running into problems and eventually had to focus my efforts into experimental work. But hopefully one day I can go back to this stuff again. Afterall, it helps us understand the reaction mechanisms better.