Documentation
Please cite the following article when referring to results from this server:
- A. Sircar,* E. Kim* & J. J. Gray,"RosettaAntibody: Antibody Variable Region Homology Modeling Server," Nucleic Acids Research 37 (Web Server Issue), W474-W479 (2009). Online | PDF (*These authors contributed equally to this work)
Details and performance of the method is described in the following article:
- A. Sivasubramanian,* A. Sircar,* S. Chaudhury & J. J. Gray,"Toward high-resolution homology modeling of antibody Fv regions and application to antibody-antigen docking," Proteins 74(2), 497-514 (2009). Online | PDF (*These authors contributed equally to this work)
Answers to frequently asked questions can be found in the documentation below. However, if you have additional questions please email us at: antibodyserver at graylab dot jhu dot edu
Flowchart
Flowchart for RosettaAntibody Fv region homology modeling protocol
Abbreviations
FV : Antibody Variable Region HFR: FV Heavy Chain Framework (VH) LFR: FV Light Chain Framework (VL) CDR: Complementarity Determining Region L1 : CDR L1 Loop L2 : CDR L2 Loop L3 : CDR L3 Loop H1 : CDR H1 Loop H2 : CDR H2 Loop H3 : CDR H3 Loop
RosettaAntibody uses the Chothia Numbering scheme for residue numbering in the model structures. The CDR loop definitions are also as given by the Chothia Numbering.
BLAST alignment terminology
Bit score
The value S' is derived from the raw alignment score S in which the statistical properties of the scoring system used have been taken into account. Because bit scores have been normalized with respect to the scoring system, they can be used to compare alignment scores from different searches.
Raw Score
The score of an alignment, S, calculated as the sum of substitution and gap scores. Substitution scores are given by a look-up table (see PAM, BLOSUM). Gap scores are typically calculated as the sum of G, the gap opening penalty and L, the gap extension penalty. For a gap of length n, the gap cost would be G+Ln. The choice of gap costs, G and L is empirical, but it is customary to choose a high value for G (10-15)and a low value for L (1-2).
BLOSUM
Blocks Substitution Matrix. A substitution matrix in which scores for each position are derived from observations of the frequencies of substitutions in blocks of local alignments in related proteins. Each matrix is tailored to a particular evolutionary distance. In the BLOSUM62 matrix, for example, the alignment from which scores were derived was created using sequences sharing no more than 62% identity. Sequences more identical than 62% are represented by a single sequence in the alignment so as to avoid over-weighting closely related family members. (Henikoff and Henikoff)
E value
Expectation value. The number of different alignents with scores equivalent to or better than S that are expected to occur in a database search by chance. The lower the E value, the more significant the score.
gap
A space introduced into an alignment to compensate for insertions and deletions in one sequence relative to another. To prevent the accumulation of too many gaps in an alignment, introduction of a gap causes the deduction of a fixed amount (the gap score) from the alignment score. Extension of the gap to encompass additional nucleotides or amino acid is also penalized in the scoring of an alignment.
Similarity
The extent to which nucleotide or protein sequences are related. The extent of similarity between two sequences can be based on percent sequence identity and/or conservation. In BLAST similarity refers to a positive matrix score.
PDB file structure (mostly from RosettaDock)
Atom coordinates
ATOM 1 N MET L 4 -15.197 35.942 2.724 1.00 0.00 ATOM 2 CA MET L 4 -14.079 35.887 3.650 1.00 0.00 ATOM 3 C MET L 4 -13.018 35.045 3.006 1.00 0.00 ATOM 4 O MET L 4 -13.330 34.005 2.426 1.00 0.00 ATOM 5 CB MET L 4 -14.492 35.189 4.942 1.00 0.00 ATOM 6 CG MET L 4 -13.330 34.946 5.913 1.00 0.00 ATOM 7 SD MET L 4 -13.650 33.611 7.170 1.00 0.00 ATOM 8 CE MET L 4 -14.586 34.554 8.356 1.00 0.00 ATOM 9 H MET L 4 -16.021 36.468 2.978 1.00 0.00 ATOM 10 HA MET L 4 -13.679 36.888 3.809 1.00 0.00 ATOM 11 1HB MET L 4 -15.243 35.818 5.416 1.00 0.00 ...
Summary
| score: | the total score using the all-atom (high-resolution) energy function (lower is better) |
| bk_tot: | total score used in the side-chain packing algorithm |
| fa_atr: | attractive portion of the lennard-jones potential (rewards close contacts) |
| fa_rep: | lennard-jones repulsive (penalizes overlaps) |
| fa_sol: | lazaridis-karplus solvation model (penalizes buried polars) |
| gsolt: | surface-area based solvation model |
| fa_dun: | internal energy of side chain rotamers as derived from dunbrack's statistics |
| fa_intrares: | intra-residue clashes |
| fa_pair: | statistics based pair term, favors salt bridges |
| hb_sc: | sidechain-sidechain and sidechain-backbone hydrogen bond energy |
| hb_srbb: | backbone-backbone hbonds close in primary sequence |
| hb_lrbb: | backbone-backbone hbonds distant in primary sequence |
ntrials: 1
%accepted: 0.00
rms: 207.75
maxsub: nalign: 212 rms: 0.19 log_Evalue: -26.86
%_nat_cont: 0.89
rms_to_start: 8.52461994E-08
score: -463.00
env: -56.50 env_weight: 0
pair: -20.18 pair_weight: 0
vdw: 1.84 vdw_weight: 0
hs: 0.00 hs_weight: 0
ss: -96.13 ss_weight: 0
sheet: 10.07 sheet_weight: 0
r-sigma: -64.89 r-sigma_weight: 0
cb: 77.17 cb_weight: 0
rg: 17.63 rg_weight: 0
co: 44.13 co_weight: 0
contact_pred: 0.00 contact_weight: 0
rama: 19.27 rama_weight: 0.2
hb_srbb: -36.73 hb_srbb_weight: 0.5
hb_lrbb: -118.27 hb_lrbb_weight: 1
bk_tot: -511.19
fa_atr: -758.29
fa_rep: 47.32
fa_sol: 372.03
fa_h2o_sol: 0.00
fa_dun: 144.08
fa_intrares: 0.20
fa_pair: -27.94
gb_elec: 0.00
fa_plane: 0.00
fa_prob: -51.94
fa_h2o: 0.00
fa_h2o_hb: 0.00
fa_ref: 35.40
gsolt: 324.70
sasa: 23135.74
hb_sc: -46.26
sasapack: -0.26
Table of residue-specific energies
| Eatr: | lennard-jones attractive |
| Erep: | lennard-jones repulsive |
| Esol: | lazaridis-karplus solvation energy |
| Eh2o_sol: | solvation using explicit water, in default mode (not used) |
| Eaa_phipsi: | prob of an aa given phi and psi (not used) |
| Edun: | rotamer internal energies |
| Eintra: | internal clashes within residues |
| Ehbnd: | total hydrogen bonding per residue |
| Epair: | pair probabilities derived from the pdb database |
| Eref: | reference energy for each amino acid |
| Egb: | generalized born solvation energy (not used) |
| Eh2o, Eh2o_hb: | energies from explicit waters (not used) |
| Ecst: | constraint energies (not used) |
| Eres: | total for that residue (lower is better) |
res aa Eatr Erep Esol Eh2o_sol Eaa Edun Eintra Ehbnd_bb Ehbnd_sc Epair Eref Egb Eh2o Eh2o_hb Ecst Eres 4 MET -3.5 0.2 1.3 0.0 0.0 1.4 0.0 0.0 0.0 0.0 0.3 0.0 0.0 0.0 0.0 -1.0 5 THR -1.8 0.1 1.1 0.0 0.1 0.1 0.0 -1.2 0.0 0.0 0.3 0.0 0.0 0.0 0.0 -1.8 6 GLN -5.2 0.5 4.3 0.0 0.1 1.2 0.0 0.0 -1.2 -0.1 1.0 0.0 0.0 0.0 0.0 -1.3 7 THR -2.2 0.3 1.3 0.0 -0.3 0.4 0.0 -1.4 0.0 -0.1 0.3 0.0 0.0 0.0 0.0 -2.3 8 PRO -2.6 0.3 1.1 0.0 0.3 0.0 0.0 0.0 -0.4 0.0 0.0 0.0 0.0 0.0 0.0 -1.3 9 LEU -1.9 0.1 1.1 0.0 -0.2 0.4 0.0 -0.8 0.0 0.0 0.1 0.0 0.0 0.0 0.0 -1.3 10 SER -1.8 0.0 1.0 0.0 0.2 0.5 0.0 0.0 0.0 -0.1 0.4 0.0 0.0 0.0 0.0 -0.6 11 LEU -3.9 0.4 0.9 0.0 0.1 0.6 0.0 -1.4 0.0 0.0 0.1 0.0 0.0 0.0 0.0 -3.3 12 PRO -1.5 0.4 0.4 0.0 -0.5 0.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 -0.8 13 VAL -2.5 0.3 0.7 0.0 0.2 0.1 0.0 -0.6 0.0 0.0 -0.3 0.0 0.0 0.0 0.0 -1.5 ...
Table of measured - expected energy
Measured - expected energy value is useful for determining how well packed a residue is. (expected energies are derived by calculating the average energies of the different amino acids with a certain number of neighbors in a large set of proteins in the pdb.) The column Elj compares the actual lennard jones energy of residue to the expected value. Well packed residues should have Elj scores new zero or negative.
| Eatr: | lennard-jones attractive |
| Erep: | lennard-jones repulsive |
| Esol: | Lazaridis-Karplus solvation |
| Eaa_phipsi: | P(aa|phi,psi) |
| Edun: | rotamer preferences from dunbrack's library |
| Eintra: | intra residue clashes |
| Ehbnd: | hydrogen bonding |
| Epair: | statistics based pair term |
| Elj: | lennard-jones total |
| Eres: | total per residue |
| SASApack: | SASApack is related to the void volume in a protein. Surface areas are computed with a 1.4 angstrom probe and 0.5 angstrom probe and the difference (ASA_0.5 - ASA_1.4) is compared to the expected difference for a particular residue type in a particular environment. A negative value is favorable and indicates that the residue is more tightly packed than is seen in average pdb files. |
energies-average(in pdb) energies, AND rsd SASA packing score res chain aa nb Eatr Erep Esol Eaa Edun Eintra Ehbnd Epair Elj Eres SASApack res_rms sasaprob 4 L MET 22 1.1 -0.2 -0.5 0.1 -1.0 -0.1 1.0 0.0 0.9 -0.1 2.68 36.44 0.363 5 L THR 13 0.8 -0.1 -0.6 0.3 -0.4 0.0 -0.6 0.1 0.7 -0.4 4.92 34.61 0.155 6 L GLN 22 -0.3 0.2 0.9 0.2 -1.0 0.0 0.3 0.1 -0.1 0.0 6.21 33.74 0.241 7 L THR 15 0.6 0.1 -0.5 -0.2 -0.2 0.0 -0.6 0.0 0.7 -0.6 -1.82 31.89 0.696 8 L PRO 13 -0.2 -0.1 0.1 1.0 -0.6 0.0 -0.1 0.0 -0.3 0.2 -3.09 30.74 0.860 9 L LEU 12 0.8 -0.1 -0.1 -0.1 -0.9 -0.1 -0.2 0.0 0.6 -1.2 3.68 31.13 0.239 10 L SER 12 0.3 -0.2 -0.4 0.3 -0.4 0.0 0.6 0.0 0.1 0.4 6.47 29.63 0.106 ...
Measured - expected energies in different environments
actual-average(in pdb) energies per residue
Eatr Erep Elj
buried 0.0 0.0 -0.1
middle 0.1 -0.1 0.1
surfac 0.2 -0.1 0.1
Difference of chi angle (not used)
decoy chi angles - starting chi angles res aa chi1 chi2 chi3 chi4 4 MET 0.00 0.00 0.00 0.00 chi_offsets 5 THR 0.00 0.00 0.00 0.00 chi_offsets 6 GLN 0.00 0.00 0.00 0.00 chi_offsets 7 THR 0.00 0.00 0.00 0.00 chi_offsets 8 PRO 0.00 0.00 0.00 0.00 chi_offsets ...
Absolute decoy chi angles
absolute decoy chi angles res aa chi1 chi2 chi3 chi4 4 MET 173.455 -159.410 -81.934 0.000 2 2 3 0 chi_absolute 5 THR -58.550 -60.000 0.000 0.000 3 0 0 0 chi_absolute 6 GLN -176.834 176.088 -22.320 0.000 2 2 3 0 chi_absolute 7 THR 59.568 180.000 0.000 0.000 1 0 0 0 chi_absolute 8 PRO 34.964 0.000 0.000 0.000 1 0 0 0 chi_absolute ...
Miscellaneous
| res: | Sequential residue number, does not match the one in pdb files |
| ss: | Secondary structure (L: loop, E: strand, H: helix) |
| phi/psi/omega: | Backbone torsion angles |
| frag | Fragments (INPT: input, -SH-: modified by shear moves, -SM-: modified by small moves) |
| rama: | Ramachandran score |
| seq: | Amino acid single letter code for that position |
res ss phi psi omega frag rama seq 94 E -67.876 151.961 -176.253 INPT -1.10 R 95 L -141.153 115.573 179.817 -SH- 1.51 W 96 E -143.038 -138.702 177.913 -SM- 4.38 G 97 L -54.728 -25.570 -173.750 -SH- -0.45 S 98 L -97.591 3.749 -149.708 -SM- -0.96 Y 99 L 47.139 -127.452 -166.108 -SM- 9.28 A 100 L -91.502 72.678 -167.819 -SH- 0.69 M 101 L -80.957 -28.905 174.097 -SM- -0.86 D 102 L -120.436 144.341 -179.498 -SH- -0.77 Y 103 L -131.996 150.399 173.932 -SM- -0.32 W 104 L -78.010 -179.619 -176.175 INPT -0.32 G ...
Command Line Options
RosettaAntibody can also be run from a command-line interface using the Rosetta package available from RosettaCommons. The server currently runs Rosetta 2.x.x using the following command-line options:
Grafting command line options
rosetta.gcc aa <4 letter pdb code> _ -antibody_modeler -nstruct 1 -quiet -l1 -l2 -l3 -h1 -h2 -graft_h3 -superimpose -deep 2 -use_pdb_numbering -ccd_closure -s <4 letter pdb code> -find_disulf -norepack_disulf
CDR H3 modeling and VL-VH optimization options
rosetta.gcc aa <4 letter pdb code> _ -s <4 letter pdb code> -nstruct 2000 -antibody_modeler -quiet -h3 -H3_filter -ex1aro -ex2aro_only -find_disulf -norepack_disulf -norepack_antibody -unboundrot -use_pdb_numbering -ccd_closure -loop_frags -build_loop -compute_hbond -max_frags 350 -relax_dock -snug_fit -freeze_h3 -flank_relax 2
*for the server: <4 letter pdb code> is "FR02"
Explanation of warnings
Warnings are displayed when the template selected for modeling a particular segment of the FV do not meet threshold values for certain filters. The filters are :
- The length of a particular segment of the template should equal the length of the respective segment of the query.
- The template structure should have a resolution better than 2.8 Ångstrom
- The minimum aligment length of a particular segment is 70% the query length of the respective segment. The factor is 55% for CDR H2 loop and 10% for CDR H3 loop.
Depending on the failure the warnings can be of the following types:
- LOW CONFIDENCE PREDICTION: This occurs when BLAST returns a sequence match, but none of the matches pass all the filters
- VERY LOW CONFIDENCE PREDICTION: This occurs when BLAST does not return any hit and the template for a particular segment is chosen based on matching the length of the query for the respective segment
- Homologous (formerly "Random") CDR H3 of same length grafted in: The warning is specific to the CDR H3 loop. Of all the CDR H3 loop structures in the RosettaAntibody database having the same length as the query CDR H3 loop, the structure with the maximum CDR H3 sequence homology to that of the query sequence is selected as the CDR H3 template. CDR H3 is grafted in only for the crude model created almost immediately on the launch of a job. The grafted CDR H3 is trimmed out prior to ab initio modeling of the CDR H3 and can not bias the modeling process in any way.
Warnings are also also displayed when ab initio modeling of CDR H3 loops result in broken loop conformations:
- Skipped X nos. low-scoring Fv models because of broken CDR-H3 loop predictions: The ten lowest scoring FV models did not have closed CDR H3 loop conformations. Outputing ten lowest scoring closed CDR H3 loop conformations was possible only after skipping X better (lower) scoring FV models with broken CDR H3 loop conformations.
- Low confidence prediction for CDR H3 loop conformation for rank Y Fv model (possibly broken CDR H3 loop): All the predicted CDR H3 loop conformations should be treated with caution. Not even ten closed CDR H3 loop conformations could be found in the 2000 FV models created by RosettaAntibody. In particular, the model ranked Y has a broken CDR H3 loop conformation.