What are the source of error of the "ab initio" methods?

I can think of the Schroedinger's equation that doens't account the relativistic speed of electrons, or the Born-Oppenheimer's approximation (separation of nuclear and electronic motions, etc.), but I am sure there are other sources. Could the 1-electron truncation be considered one? And is there any other?