Homology modeling and HADDOCK3 docking of adenylate kinases.
gEDES-HM transfers a conformational ensemble generated by enhanced-sampling simulation of one protein to its homologues by multi-template homology modeling. This repository holds the scripts, modeling inputs, docking results and analysis data for its validation on 22 adenylate kinase homologues.
Paper: Kurt et al., J. Chem. Inf. Model. 2026, 66, 13586-13596.
The ensemble itself comes from the gEDES protocol: vargiulab/gEDES, Basciu et al., J. Chem. Inf. Model. 2025.
The complete archive is on Zenodo: 10.5281/zenodo.21534586. This repository has the same layout, without the structure ensembles, which are too large for git (710 MB).
| GitHub | Zenodo | |
|---|---|---|
scripts/, analysis/, docking/ |
yes | yes |
proteins_structural_models/: MODELLER drivers, alignments, RMSD wrappers |
yes | yes |
proteins_structural_models/: templates, models, references (*.pdb.gz) |
no | yes |
To run the steps that need the structures, download
proteins_structural_models.zip from Zenodo and unpack it at the top of this
repository:
curl -L -o proteins_structural_models.zip https://zenodo.org/records/21534586/files/proteins_structural_models.zip
unzip -n proteins_structural_models.zipThese work from this repository alone:
cd analysis/tableS6_check && bash check_tableS6.sh
cd analysis/statistics && python3 gEDES_SI_statistics.pyREADME.md
proteins_structural_models/ modeling inputs, one folder per system, 1 to 22
gEDES/ apo/ holo/
docking/ HADDOCK3 results, all four receptor conditions
analysis/ computed data behind the reported numbers
scripts/ workflow scripts
Each subdirectory has its own README with the detail: docking/README.md for
the file naming and the columns of the ranking tables, analysis/README.md for
the data tables and how to regenerate them, scripts/README.md for the
workflow.
| Label | Where | Receptors per target |
|---|---|---|
| HM_gEDES | gEDES/, docking/<n>_* |
101 homology models from the gEDES representatives, DOPE-selected |
| HM_4AKE(apo) | apo/, docking/APO_* |
10 models from the single apo template, PDB 4AKE |
| HM_1AKE(holo) | holo/, docking/HOLO-1ake_* |
10 models from the single holo template, PDB 1AKE |
| HOLO_expt | docking/redocking_HOLO-expt_* |
1, the target's own experimental holo receptor, ligand removed |
The first two were in the original submission. HM_1AKE(holo) and HOLO_expt
were added in revision as controls: the first bounds what the transferred
ensemble can achieve given an experimental bound structure of the reference
protein, which a prospective application would not have; the second is the
redocking ceiling and the noise floor of the RMSD metrics.
Files are keyed by system number, 1 to 22. Four targets carry two ligands each
and were docked as separate complexes, so a ligand suffix distinguishes them:
system 2 _ADP / _AMP, system 9 _ADP / _AMP, system 15 _ADP182 /
_ADP183, system 21 _ADP214 / _ADP215. That is why the 22 systems give 26
complexes throughout.
| System | Target | System | Target |
|---|---|---|---|
| 1 | A0A0J9X1X4 | 12 | Q9UIJ7 |
| 2 | Q6LTE1 | 13 | Q04ML5 |
| 3 | P84139 | 14 | P27144 |
| 4 | P16304 | 15 | P9WKF5 |
| 5 | P07170 | 16 | A0A2R2JFU6 |
| 6 | P27142 | 17 | Q68EH2 |
| 7 | Q6B341 | 18 | M4AA20 |
| 8 | O66490 | 19 | A0A2R2JFU5 |
| 9 | Q8IJV6 | 20 | A0A452CSM2 |
| 10 | P43188 | 21 | 5G3Y (PDB) |
| 11 | P08760 | 22 | 5G3Z (PDB) |
UniProt accessions unless marked as PDB. Systems 21 and 22 are synthetic constructs without UniProt entries.
- Structures are PDB format, many gzipped;
.tsvfiles are tab separated. - The binding-site definition for a given target is the
BS_RESIDUESvalue in itsrun_calc_rmsd-cpptraj*.sh, and is the same at the modeling and the docking stage. - Nothing under
analysis/contains an absolute path. SetADK_ARCHIVE_ROOTonly if you separate it from the rest of the archive.
If you use gEDES-HM, please cite the paper:
Kurt H, Athar M, Basciu A, Reys V, Bonvin AMJJ, Vargiu AV. Mapping Functional Transitions across Protein Homologues from a Single Simulation: Application of gEDES-HM to Adenylate Kinases. J. Chem. Inf. Model. 2026, 66 (20), 13586-13596. https://doi.org/10.1021/acs.jcim.6c02521
The full data archive:
Kurt H, Athar M, Vargiu A. Mapping functional transitions across protein homologues from a single simulation: application of gEDES-HM to adenylate kinases. Zenodo, 2026. https://doi.org/10.5281/zenodo.21534586
CC-BY-4.0, the same as the Zenodo archive. See LICENSE.
