######################################################
# --- KORP v1.22 release - November 26th, 2018 --- #
# --- http://chaconlab.org/modeling/korp --- #
######################################################
----------------------------------------------
KORP (Knowledge based ORientational Potential)
----------------------------------------------
Pre-compiled 64-bit LINUX binaries (v1.22) are available in the bin/ directory.
We recommend adding binaries path to your PATH environment variable or just
copy them to your favorite "bin" directory.
Please, choose the appropriate release for your system:
[Program] Compiler Libraries Linkage Version
------------------------------------------------------------------------
korpe* Intel icpc (v16.0.1) - static 1.22
korpe_gcc GNU gcc (v4.6.3) - static 1.22
* Intel compiled binary is the fastest alternative (10-20% faster).
##################
# KORPE TUTORIAL #
##################
The basic instructions to compute the KORP energies with "korpe" tool are
described here (for both loops and complete proteins).
The released korp6Dv1.bin energy map contains the information from ALL the
36851 protein chains of our training set (listed in PISCES_id90_r3A.txt file).
PLEASE, BE SURE TO PROVIDE A VALID PATH TO korp6Dv1.bin ENERGY MAP.
The PDB protein chains with resolution better than 3.0 A, R-factor better
than 1.0, and 90% maximum sequence identity were obtained from PISCES server
(http://dunbrack.fccc.edu/Guoli/pisces_download.php) on October, 4th 2018.
More info about KORP is available at http://chaconlab.org/modeling/korp
---------------
Loop modeling
---------------
The KORP energy computation for loop ensembles is fast and easy with "korpe".
> Single run
------------
For a single case run, just provide a PDB file (filaname.pdb) of the protein
and a Multi-PDB file with all the loops to be assessed (multi_loops.pdb).
#> korpe filename.pdb --loops multi_loops.pdb --score_file path_to_map/korp6Dv1.bin -o basename
In the output basename.txt file you will find a table with the KORP energies.
For example, in the directory rcd6/ you can run:
#> korpe 3CXM.pdb --loops 3CXM_closed.pdb --score_file ../korp6Dv1.bin -o 3CXM_result
and check the 3FHD_result.txt output file.
> Multiple run
--------------
For a multiple cases run, use a plain text file with the base names of the
protein PDBs, i.e. a list of the target PDBs file names without the .pdb
extension (filename_list.txt), and the common suffix of the loops Multi-PDB files
(_loops.pdb) instead.
#> korpe filename_list.txt --loops _loops.pdb --score_file path_to_map/korp6Dv1.bin -o suffix
For each XXX entry in the filename_list.txt you will find a file name
XXX_suffix.txt with the corresponding a table of KORP energies.
For example, in the directory rcd6/ you can run:
#> korpe rcd6_ids.txt --loops _closed.pdb --score_file ../korp6Dv1.bin -o output
*TIPS
-----
- The presence of one N-terminal and one C-terminal residue (anchors)
in the Multi-PDB file (or the _loops.pdb files) is mandatory.
- Both the protein PDB and the loops Multi-PDB must be numerated
consistently.
- KORP energy is side-chain independent, it only requires the coordinates
of the N, CA, and C mainchain atoms.
- If benchmarking, add the --rmsd flag and provide a complete PDB (with native
loop) to obtain complete KORP energy statistics in the output_score.txt file.
For example, use these comands for single and multiple runs, respectively, in
the rcd6/ directory:
#> korpe 3CXM.pdb --loops 3CXM_closed.pdb --score_file ../korp6Dv1.bin -o 3CXM_result --rmsd
#> korpe korp6_ids.txt --loops _closed.pdb --score_file ../korp6Dv1.bin -o output --rmsd
------------------
Protein modeling
------------------
The KORP energy computation for protein models is fast and easy with "korpe",
> Single case run
-----------------
For a single case run, just provide a PDB file of the protein and the
corresponding energy map. For example, in directory CASP12DCsel20/ run:
#> korpe T0902D1_s432m2.pdb --score_file ../korp6Dv1.bin
The computed energy will be prompted to screen.
> Multiple run
--------------
For a multiple cases run, just input a plain text file with the base
names of the PDBs, i.e. a list of PDB file names without the .pdb
extension. For example, in directory CASP12DCsel20/ run:
#> korpe casp12dcsel20.txt --score_file ../korp6Dv1.bin -o korp6D
*TIPS
-----
- Use a correct residue numeration in the PDB, it may be important for
bonding/non-bonding contacts discrimination.
- KORP energy is side-chain independent, it only requires the coordinates
of the N, CA, and C mainchain atoms.
##################
# KORPE timmings #
##################
Benchmark Targets Structures* Time^
[#] [#] [s]
----------------------------------------
rcd6 100 100100 33
rcd8 100 100100 42
rcd10 100 100100 53
rcd12 100 100100 59
casp10all 78 25092 163
casp11all 58 17507 116
casp12all 32 10056 62
casp10best150 78 9232 67
casp11best150 55 7680 54
casp12best150 28 3745 28
casp10sel20 73 1333 11
casp11sel20 46 914 7
casp12sel20 23 470 5
rosetta41 41 4599 17
itasser56 56 24650 74
3DRobot 200 60200 371
rosetta3DR41 41 4141 17
itasser3DR56 56 22456 83
----------------------------------------
* Native structures included.
^ Computation time obtained using the Intel compiled version of korpe and
the provided all-heavy-atoms benchmarks in an old Intel(R) Core(TM)
i7-950 (3.07GHz) workstation. Note that korpe would be around two times
faster if you used just the N, CA, C atom coordinates as input (instead
of the all-heavy-atoms benchmarks).
----------
REFERENCES
----------
Please, cite our work if our tools or benchmarks become useful for your
research.
> KORP energy and benchmarks
----------------------------
Lopez-Blanco JR and Chacon P (2018). KORP: Knowledge-based 6D
potential for protein structure prediction. (to be published)
> Improved RCD method
---------------------
Lopez-Blanco JR, Canosa-Valls AJ, Li Y, and Chacon P (2016). RCD+: Fast loop
modeling server. NAR (DOI: 10.1093/nar/gkw395).
> Original RCD method
---------------------
Chys P and Chacon P (2013). Random coordinate descent with spinor-matrices and
geometric filters for efficient loop closure. J. Chem. Theory Comput.
9:1821-1829.
-------
CONTACT
-------
Please, feel free to contact with us!
(suggestions or bug reports are welcome)
Jose Ramon Lopez-Blanco (PhD.)
This email address is being protected from spambots. You need JavaScript enabled to view it.
Pablo Chacon (PhD.)
This email address is being protected from spambots. You need JavaScript enabled to view it.
