For each of the targets below you have to:

  Go to the CASP3 site and bring ALL the models that start with:
 T00nnFFmmm_1
 where nn is the target number taken from the list below,
 FF is either "TS" or "AL",
 mmm is the group name doing the prediction,

 Note that there are some models of the form 
T00nnFFmmm_1_1 T00nnFFmmm_1_2. These correspond to split models. You have
to evaluate each separately and then add the results of the GL to give
a single number.

To evaluate the models you will need the PDB code of the real structure.
You can find all the available PDB's in the target list of the CASP3 page.
If a given PDB is not available there (but appears in my list) then
skip it.

We have two different tables to build. 
The list for table I is:
81 44 85 83 54 53 51 63 79 78 80 46 71 67 43 59 72 77 61 75

The list for table II is:
47 58 60 64 76 48  74 82 55 49 62 57 73 70 68

Note that the PDB files of some of the targets need be split into different targets:
76: residues 1-71 will be 761   and residues 72-140 will be 762.
71: residues 1-121  will be 711 and residues 122-238 will be 712.
64: residues 1-68 only.
74: residues 121-197 only.
63: residues 71-138 only.
83: residues 1-105 only.

The following target needs be split into two, but report only the best GL result
from each splitted component: 
79: residues 1-75  OR 55-124.