IUPAC definition of chi1 for VAL



 I am having a bit of trouble "grokking" the IUPAC definitions of the
 chi
 angles (side-chain dihedral angles) for amino acids.  I _think_ I have
 it right, but I'd like someone who thinks he/she knows for sure to
 confirm my reading of the definitions.  The only cases I'm uncertain about
 is where there are two "identical" (but non-symmetrical) 4th atoms in
 the
 torsion, as in chi1 of VAL or chi2 of LEU;  in this case the IUPAC dictates
 which of the atoms is to be labeled the "main" branch.  If this naming
 convention is followed, then, as I understand it, chi1 of VAL would be
 defined by N-CA-CB-CG1, and chi2 of LEU would be CA-CB-CG-CD1.  In Brookhaven
 files, most VALs and LEUs have their terminal branches numbered consistently
 as shown below; however, there are a number of exceptions.
 Anyway, here is my understanding of how the angles and corresponding
 rotamers should be named for VAL, using Newman projections looking down
 the CB-CA bond, with CB closer to you.  I also have the chirality of the
 alpha-carbon right, I believe/hope.
                  O
                 /
                /
               C                            This is the t (trans) conformation.
               |
               |                            Chi1, the dihedral angle specified
       CG1     |
          \    |                             by N-CA-CB-CG1, is ~180 deg.
           \   |
            \-----)                         It would seem to make sense that
           ( \     )
          (   \     )                        this rotamer is lowest in energy,
          (    *    )
          (    |    )                        since it has the fewest 1-4 heavy-
           (   |   )
            (--|--)                          atom interactions, and in fact,
 	      |   \
 	      |    \                        Ponder & Richards state that the
 	      |     \
 	      CG2    NH                     trans conformation is observed
                                             for VAL in proteins about 67% of
                                             the time.
                  O
                 /
                /
               C
               |
               |      GG2                   This is the - (gauche -) conf.
               |     /
               |    /                       Chi1 is ~ -60 deg.
               |   /
            (-----/
           (     / )
          (     /   )
          (    *    )
          (    |    )
           (   |   )
            (--|--)
 	      |   \
 	      |    \
 	      |     \
 	      CG1    NH
                  O
                 /
                /
               C
               |
         CG2   |      GG1                   This is the + (gauche +) conf.
         \     |     /
          \    |    /                       Chi1 is ~ +60 deg.
           \   |   /
            \-----/
           ( \   / )
          (   \ /   )
          (    *    )
          (         )
           (       )
            (-----)
 	          \
 	           \
 	            \
 	             NH
 So, someone out there please tell me if I am right.  Incidentally, in
 the molecular modeling package Insight, if I pick "Torsion" from the
 "Transform" menu, and, in two-atom mode, define a torsion about CA-CB,
 the program uses the atoms C-CA-CB-CG1, not what I believe is the IUPAC
 definition of chi1, as the dihedral angle.
 	-P.
 ************************f*u*cn*rd*ths*u*cn*gt*a*gd*jb*************************
 Peter S. Shenkin, Department of Chemistry, Barnard College, New York, NY 10027
 (212)854-1418    shenkin-0at0-avogadro.barnard.columbia.edu
 shenkin-0at0-cunixc.BITNET
 ********** "I've got algorithm -- who could ask for anything more?"
 **********