Helix-helix packing angle preferences for finite helix axes

Recently, James Bowie addressed the question of how to normalize correctly the distribution of observed helix-helix packing angles in proteins (Bowie, Nature Struct, Biol, 4:915-917, 1997). A hitherto unrealized yet significa nt bias toward crossed packing angles was revealed. However, the derived ra ndom reference distribution of packing angles requires that helices have to be assumed as infinite in length. Here, we complement Bowie's analysis by consideration of the more realistic case where helices are of finite length . As a result, the statistical bias toward near perpendicular packings appe ars to be even stronger, (C) 1998 Wiley-Liss, Inc.