W:MCSCF GAMESS-US input?

Hi!
 I have a question regarding MCSCF runs using GAMESS-US. I am familiar with UHF
 and DFT runs, but MCSCF is new.
 The molecule is a metalloporphyrin, ClFe(III)porphyrinato. I have run a
 preliminary UHF computation using SBKJC-ECP for FeIII and a 6-31G* for the rest
 of the molecule. A "modest" zero order determinant to start the MCSCF
 run.
 The active space would consist of the 6 cartesian d orbitals and the "a1
 HOMO-1, a2 HOMO, and e LUMO" centered on the porphyrin ring. This follows
 the "four orbital model" of Gouterman.
 By inspecting the output and looking at the orbitals I found that the
 6 cartesian d orbitals are numbers 45-49,51 in the list of alpha orbitals while
 the a1,a2, and e orbitals are 93-96 in the alpha orbitals and 88-91 in the beta
 orbitals.
 My problem is in setting the options to define this active space.
 Looking at the manual it looks like I need to do the following:
 1) Set RUNTYP=MCSCF in CONTRL
 2) Set GUESS GUESS=MOREAD NORB=426  to read all the orbitals, alpha and beta
 using  the VEC namelist
 3) Set DET NCORE=87 NACT=14 NELS=9 GROUP=C2V ISTSYM=1 NSTATE=1
 According to the manual, the Sz will be picked up from the multiplicity which 6
 (2S+1), so Sz=5/2.
 I have 94 alpha electrons and 89 beta electrons (183 total). This means that 9
 electrons are in the active space (5 d electrons and 4 from the HOMO-1 and
 HOMO). The remainder will be "inactive" and evenly divided between
 alpha and beta orbitals (87 * 2 = 174). The active space has a total of 14
 orbitals (6 cartesian d orbitals [alpha 45-49,51] and 8 orbitals from the
 porphyrin ring [alpha and beta: a1 HOMO-1, a2 HOMO, and e LUMO]). GAMESS only
 uses Abelian groups for MCSCF so must drop from C4v to C2v. The last two options
 request only one state of A1 symmetry.
 Question 1)
 It is not clear to me how the program will pick up the correct number of alpha
 and beta orbitals with the above input. It seems that something is
 missing to force the program to split the active orbitals into 10 alpha and 4
 beta orbitals. (This may be easier with the 5 spherical d orbitals. In this case
 NACT=13. By making Sz=5/2 this will pick 5 alpha orbitals for the active space
 and split the rest evenly between alphas and betas.)
 Does any one have any suggestions or can shed light on this problem? Or the
 answer is run with ISPH=1 and get the 5 spherical d orbitals!
 Question 2)
 Even if I solve question 1, I will have to reorder the orbitals. In g03 you can
 use guess alter combo. In GAMESS-US you can use NORDER=1 under GUESS and then
 enter IORDER and JORDER. My problem is what is the proper format for entering
 IORDER and JORDER for this case.
 The idea would be to reorder the input orbitals so that the top 10 alpha
 orbitals correspond to the 6 cartesian d orbitals, the a1 HOMO-1, a2 HOMO, and
 the e LUMO while doing the same for the beta orbitals so that the top 4 are the
 a1 HOMO-1, a2 HOMO, and e LUMO.
 I've used NORDER=1 and IORDER(45)=45,88 to swap 45 and 88 in the list of alpha
 orbitals, but the run aborts complaining that I have not given some of the
 input. It is like once I start entering values for orbital 45 it expects input
 for all the remaining orbitals up to 426!.
 Any suggestions here?
 Thanks!
 Gustavo Mercier MD,PhD
 Baylor University Medical Center
 Dallas, TX
 gamercier -()- yahoo.com
 gustavom -()- baylorhealth.edu