I accidently read this page.. http://www.majorgroove.org/questions/93/leu-val-rotamers
It explains a nightmare scenario which has been bugging me for some time asw I tried to make a beautifully defined VAL fit the density in COOT..
If by some accident your VAL or LEU have the "wrong" naming this is acceptable to REFMAC who assigns the chirality at CB or CG as "both". However it is not acceptable to COOT
In fact, the default monomer library used for COOT is different to that used by REFMAC in this detail
I hope the distributed libraries will get back in step asap. Grrrrrrrr Eleanor