LOMETS v1.5
                                                             Last updated: 05/02/2009
                                                             email to: zhanglab@ku.edu

LOMETS (LOcal MEta-Threading-Server) is a local consensus method for protein structure prediction. It collects 9 different individual servers   and re-ranks
them by using consensus method. The most important advantage of LOMETS is that all the inidivual servers are run locally.  Therefore  the  results
are generated locally by individual servers. This ensures the reliability and fastness of meta-server. Another advantage of LOMETS is that it gives 
additional restraint files that can be further used in ab initio modeling such as TASSER.

The 9 component servers of LOMETS are
   Method                 Z-score cutoff
1. MUSTER         [1]     5.8
2. HHsearch v1.0  [2]     13.0
3. HHsearch v1.5a [2]     9.5
4. HHsearch v1.5b [2]     11.5
5. SPARKS         [3]     7.0
6. SP3            [4]     7.0
7. PROSPECT2      [5]     3.2
8. PPA-I          [6]     9.3
9. FUGUE          [7]     7.0

For LOMETS and its 9 component servers, we give top 10 models with corresponding alignment and threading results.
1. Rank: Top 1 to Top 10 models.
2. Template: the template identified by LOMETS.
3. Align_length: The length of aligned region in the threading results of LOMETS.
4. Coverage: Align_length/target length
5. Zscore=(raw_score-mean_score)/std_score
6. Seq_id: Sequence identity by threading results
7. Confidence: If Z-score > Z-score cutoff, the corresponding template is predicted with high confidence. Otherwise, it is a template with low confidence.
8. Program : threading program
9. Target-template alignment: The alignment files describe the sequence alignments of templates to sequence after threading.
10. 3-D models from threading alignments: The threading files describe the aligned region between templates and target sequences with the corresponding
   3-D coordinates of templates. Columns 4 and 5 are the residue names and residue orders of C alpha atoms of targe sequences, respectively. Columms 9
   and 10 are the residue orders and residue names of C alpha atoms of aligned templates, respectively.
11. Full-length models by MODELLER: The final models are generated by MODELLER v8.2.


For LOMETS, there are 4 additional restraint files: 
1. Sidechain contact file describes the predicted sidechain contacts between pairs of residues. First line describes the total number of pairs 
of predicted sidechain contacts. Columns 1 and 2  are residue orders of predicted pairs. Column 3 is the percentage of servers predicting the 
sidechain contacts.If all the component servers predict the contacts, this value is 1.00.
2. C alpha atom contact file describes the predicted C alpha atom contacts between pairs of residues. First line describes the total number of 
pairs of predicted C alpha contacts. Columns 1 and 2  are residue orders of predicted pairs. Column 3 is the percentage of servers predicting 
the C alpha contacts.
3. Short-range distance file describes the predicted short-range distances of C alpha atoms. First line describes the total number of short-range
C alpha atom pairs. Columns 1 and 2  are residue  orders of predicted  pairs.  Column 3   is the percentage of servers predicting the short-range 
distances. Column 4 is the average distance between C alpha atom pairs predicted by different servers. Column 5 is the standard-deviation distance
 between C alpha atom pairs by different servers. 
4. Long-range distance file describes the predicted long-range distances of C alpha atoms.First line describes the total number of long-range C 
alpha atom pairs. Columns 1 and 2  are residue orders of predicted pairs.Column 3 is the distance between C alpha atom pairs. There are up to 4 
values of distances for each pair.

References:
[1] Sitao Wu, Yang Zhang.(2008) MUSTER: Improving protein sequence profile-profile alignments by using multiple sources of structure information. Proteins, vol 72, 547-556. 
[2] Soding, J. (2005) Protein homology detection by HMM-HMM comparison. Bioinformatics (Oxford, England), 21, 951-960.
[3] Zhou, H. and Zhou, Y. (2004) Single-body residue-level knowledge-based energy score combined with sequence-profile and secondary structure information for fold recognition. Proteins, 55, 1005-1013.
[4] Zhou, H. and Zhou, Y. (2005) Fold recognition by combining sequence profiles derived from evolution and from depth-dependent structural alignment of fragments. Proteins, 58, 321-328.
[5] Xu, Y. and Xu, D. (2000) Protein threading using PROSPECT: design and evaluation. Proteins, 40, 343-354.
[6] Sitao Wu, Yang Zhang. (2007) LOMETS: A local meta-threading-server for protein structure prediction. Nucleic Acids Research, vol 35, 3375-3382. 
[7] Shi, J., Blundell, T.L. and Mizuguchi, K. (2001) FUGUE: sequence-structure homology recognition using environment-specific substitution tables and structure-dependent gap penalties. J Mol Biol, 310, 243-257.