UnibindMD: An Integrated AI and Molecular Dynamics Platform for De Novo Peptide Binder Design
- 1ShenZhen UnibindMD Computation Biotech Co., Ltd.
Abstract
UnibindMD is an integrated computational platform that turns a protein target into validated de novo peptide-binder leads by chaining five orthogonal stages: deep-learning binder design (BindCraft), independent structure prediction and confidence assessment (AlphaFold3-class / Boltz-2), binding-affinity prediction (PRODIGY), developability profiling (ADMET-AI), and all-atom molecular-dynamics validation (GROMACS/CHARMM36m). The platform targets a persistent weakness of single-method design, namely that confidence metrics such as ipTM and pLDDT are self-consistent but not independent, so a design predicted to be confident may still be physically unstable. We demonstrate UnibindMD on three Andes virus (ANDV) intervention points, a setting of urgent need given a case fatality rate approaching 40%, a 2026 MV Hondius cruise-ship outbreak (13 cases, 3 deaths), and no approved therapeutic: the Gn-Gc glycoprotein complex (viral entry), the L-protein endonuclease domain (cap-snatching; PDB 6Q99), and the host VEGFR2 kinase (vascular permeability; PDB 2XIR). Across nine BindCraft campaigns (4,475 trajectories) the platform produced 464 accepted binders (acceptance 2.3-67.6%), with top candidates reaching ipTM = 0.94 and pLDDT = 0.94. The three selected leads were predicted by PRODIGY to bind with G = -9.1 to -11.1 kcal mol-1 (Kd 8 nM-0.2 M), and all three remained stably bound over 100 ns of unrestrained molecular dynamics (receptor CRMSD [≤] 2.6 [A], retained peptide fold, [≥] 99.5% of frames preserving more than half of the initial interface contacts, and 75-82% helical content). ADMET-AI flagged the expected poor oral bioavailability of large polar peptides and no PAINS alerts. By converting single-prediction design into a cross-validated, dynamics-tested workflow, UnibindMD provides a rapid, scalable, and generalizable path from any protein target to validated peptide-binder candidates.
Related
- PP-MAPS: Dynamic Pharmacophore Signatures of Protein–Peptide Interfaces from Molecular Dynamics Trajectories
- Accelerating Molecular Dynamics with Hydrogen Mass Repartitioning to Prioritize Immunogenic Neoantigens for Cancer Vaccines through Peptide HLA Binding Free Energy Calculations
- CasanovoGUI: a cross-platform desktop application for deep learning-based de novo peptide sequencing with Casanovo
- ToxINsight: An Integrated Web Platform for Sequence, Structure, and Ligand Binding Analysis of Snake Venom Toxins
- VOC·BIO Library: An Integrated Platform for Published Aphid GC-MS Evidence, Biological Contexts, Cheminformatics and Exploratory OBP Docking
BibTeX
@misc{wu2026unibindmd,
title = {{UnibindMD: An Integrated AI and Molecular Dynamics Platform for De Novo Peptide Binder Design}},
author = {X. Wu and A. Huang and C. Fu and Z. Wang and W. Fu},
howpublished = {bioRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.20.760329},
url = {https://www.biorxiv.org/content/10.64898/2026.10.20.760329v1},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - UnibindMD: An Integrated AI and Molecular Dynamics Platform for De Novo Peptide Binder Design
AU - X. Wu
AU - A. Huang
AU - C. Fu
AU - Z. Wang
AU - W. Fu
PY - 2026
DA - 2026/10/20
PB - bioRxiv
DO - 10.64898/2026.10.20.760329
UR - https://www.biorxiv.org/content/10.64898/2026.10.20.760329v1
AB - UnibindMD is an integrated computational platform that turns a protein target into validated de novo peptide-binder leads by chaining five orthogonal stages: deep-learning binder design (BindCraft), independent structure prediction and confidence assessment (AlphaFold3-class / Boltz-2), binding-affinity prediction (PRODIGY), developability profiling (ADMET-AI), and all-atom molecular-dynamics validation (GROMACS/CHARMM36m). The platform targets a persistent weakness of single-method design, namely that confidence metrics such as ipTM and pLDDT are self-consistent but not independent, so a design predicted to be confident may still be physically unstable. We demonstrate UnibindMD on three Andes virus (ANDV) intervention points, a setting of urgent need given a case fatality rate approaching 40%, a 2026 MV Hondius cruise-ship outbreak (13 cases, 3 deaths), and no approved therapeutic: the Gn-Gc glycoprotein complex (viral entry), the L-protein endonuclease domain (cap-snatching; PDB 6Q99), and the host VEGFR2 kinase (vascular permeability; PDB 2XIR). Across nine BindCraft campaigns (4,475 trajectories) the platform produced 464 accepted binders (acceptance 2.3-67.6%), with top candidates reaching ipTM = 0.94 and pLDDT = 0.94. The three selected leads were predicted by PRODIGY to bind with G = -9.1 to -11.1 kcal mol-1 (Kd 8 nM-0.2 M), and all three remained stably bound over 100 ns of unrestrained molecular dynamics (receptor CRMSD [≤] 2.6 [A], retained peptide fold, [≥] 99.5% of frames preserving more than half of the initial interface contacts, and 75-82% helical content). ADMET-AI flagged the expected poor oral bioavailability of large polar peptides and no PAINS alerts. By converting single-prediction design into a cross-validated, dynamics-tested workflow, UnibindMD provides a rapid, scalable, and generalizable path from any protein target to validated peptide-binder candidates.
ER -