I interpret your question to mean that you have a long loop which you think contains an 8-residue helix. You have 1000 models and want to know which, if any, have the secondary structure you need.
The simplest way to do this would be to use a DSSP tool to assign secondary structure for the protein and then search that database for structures with H in the desired region. For example, concatenate your DSSP assignments (labeled with file names) into one big file and write a script to sort the entries based on the number of “H” within the right range.
Rosetta is definitely capable of doing DSSP assignments, but I’m not sure that we have a boxed application that just takes input and reports DSSP. Writing yourself a quick script to do this in PyRosetta would be very simple; in pseudocode (this is not real code, but to give you an idea):
for (all PDBS)
load pdb into pose
print pose.dssp
You could combine it with a sorting script:
for (all PDBs):
load pdb into pose
string dssp = pose.dssp
int number_H
for thisposition in positions_in_loop:
if dssp[thisposition] == “H”:
number_H++
print dssp, number_H, posename
then sort the results on number_H.