Processing...
Clean a PDB File FREE
Results
Change Log
Downloads
Geometry Validation FREE
BioPython-based geometry checks — no external tools, results in ~2–5 seconds.
- Accepts: Standard PDB format (.pdb, .ent)
- Checks: Backbone completeness · Ramachandran outliers · Steric clashes · B-factor disorder · Chain gaps
- Skips: Ligands, water, non-standard residues (HETATM) — protein backbone only
- Multi-model NMR: Only the first MODEL is analyzed
- Does not accept: mmCIF / .cif format
Merge Structures for Docking FREE
Combine two PDB files into one — chain IDs are deduplicated and structures are spatially separated so docking software can find the binding pose.
- Use case: Antibody (H & L chains) + Antigen → single file for AutoDock, HADDOCK, ClusPro
- Chain conflicts: Duplicate chain IDs in Structure 2 are automatically renamed
- Separation: Structure 2 is translated along X by the specified gap (edge-to-edge)
- Multi-model NMR: Only the first MODEL from each file is used
- Need to merge one antibody with many antigens? Use Batch Merge (paid).
Prodigy FREE
Integration in ProgressProtein-protein and protein-small molecule binding affinity prediction from the Bonvin Lab (Utrecht). Upload a complex PDB, specify the interacting chains — returns predicted ΔG (kcal/mol) and Kd at 25°C.
What Prodigy calculates
- Binding free energy (ΔG) from interface contacts
- Dissociation constant (Kd) at 25°C
- Interfacial contact statistics (polar, apolar, charged)
- Works on antibody-antigen, protein-protein, protein-peptide complexes
Vangone A & Bonvin AMJJ. Contacts-based prediction of binding affinity in protein–protein complexes. eLife 4:e07454 (2015). DOI:10.7554/eLife.07454
Xue LC et al. PRODIGY: a web-server for predicting the binding affinity of protein–protein complexes. Bioinformatics 32:3676–3678 (2016). DOI:10.1093/bioinformatics/btw514
BUDEAlaScan FREE
Integration in ProgressComputational alanine scanning mutagenesis from the Woolfson Lab (Bristol/Crick). Identifies key interface residues by mutating each to alanine and computing the change in binding free energy (ΔΔG).
What BUDEAlaScan calculates
- ΔΔG for each residue at the protein-protein interface
- Hot-spot residues (large positive ΔΔG = critical for binding)
- Ranked residue importance for interface engineering
- Compatible with antibody-antigen and general PPI complexes
TAP FREE
Integration in ProgressTherapeutic Antibody Profiler (Oxford). Assesses antibody developability: humanness, post-translational modification (PTM) sites, structural liabilities, and biophysical properties of the CDR loops.
What TAP assesses
- CDR loop length and structural liability flags
- Sequence-based PTM site prediction (deamidation, oxidation, isomerization)
- Germline humanness score
- Biophysical property estimates (hydrophobicity, charge)
TAP takes antibody variable domain sequences (not a PDB file). Enter your VH and VL sequences below, then open TAP to paste them in.
EvoEF FREE
Integration in ProgressEvoEF2 is an energy function for protein design and stability/binding free energy calculations (Huang lab, Westlake University). Supports mutation scanning, interface ΔΔG computation, and protein stability evaluation.
What EvoEF2 calculates
- Protein stability ΔG from physical energy terms
- Binding free energy ΔG for protein complexes
- ΔΔG for point mutations (stability and binding effects)
- Side chain repacking and energy minimization
EvoEF2 proxy coming soon.
ABodyBuilder3 FREE
ABodyBuilder3 (Exscientia) predicts the 3D Fv structure of antibodies from paired VH and VL sequences — no template required. State-of-the-art CDR H3 loop modelling runs locally on this server.
What ABB3 provides
- Full Fv structure prediction from VH/VL sequences
- State-of-the-art CDR H3 loop modelling
- Confidence scores per residue
- Output ready for docking or MD prep
Abanades B et al. ABodyBuilder3: improved and scalable antibody structure predictions. Bioinformatics 40(10):btae576 (2024). DOI:10.1093/bioinformatics/btae576
IgFold FREE
IgFold (Gray Lab, JHU) predicts antibody and nanobody structures directly from sequence using a graph transformer trained on 558M natural antibody sequences. VL is optional — omit it for single-domain antibodies and nanobodies.
What IgFold provides
- Fast Fv structure prediction (~25 s on CPU)
- Nanobody / VH-only support (leave VL blank)
- Chothia renumbering applied automatically
- Per-residue confidence in B-factor column
Ruffolo JA et al. Fast, accurate antibody structure prediction from deep learning on massive set of natural antibodies. Nature Communications 14:2389 (2023). DOI:10.1038/s41467-023-38063-x
Batch Process Multiple PDB Files PAID
Batch Results
File Results
| Filename | Status | Safe | Gaps | Time | Actions |
|---|
Batch Summary
Downloads
Batch Validate PAID
Run BioPython geometry validation across a batch of PDB files. Produces a summary table with quality scores, flagged issues, and per-file downloadable reports.
Batch Relax & Validate PAID
Energy-minimise each structure with OpenMM (AMBER14 + implicit solvent), then run full geometry validation before and after. Ideal for ABB3 / IgFold predictions before docking. Allow 1–3 minutes per file on CPU. Maximum 5 files per submission.
Cost: 0.05 credits per file.
Batch Merge PAID
One-to-many merging: upload a single anchor structure and a batch of partner structures. Each partner is merged with the anchor and returned as a ZIP — ideal for screening many antibodies against one antigen (or vice versa).
Modes
- Antigen anchor — 1 antigen + N antibodies → N merged complexes
- Antibody anchor — 1 antibody + N antigens → N merged complexes
Batch ABB3 PAID
Predict up to 5 antibody Fv structures in a single submission. Provide paired VH and VL sequences for each antibody — the service returns a ZIP of all predicted PDB files.
Preset Pipelines PAID
Chain multiple tools automatically in a single submission. Upload once — the pipeline handles each step and returns all intermediate and final outputs.
Antibody Modeling & CDR QC FREE
Predict antibody Fv structure from VH/VL sequences, then run CDR-level geometry quality control — before docking or MD prep. Returns per-CDR Ramachandran outlier counts, heavy-atom clashscore, and a docking readiness assessment based on evidence-based thresholds from a 47-pair SAbDab benchmark.
⏱ ~1 min (ABB3) · per-CDR Rama & clashscore · Pattern A/B docking gates
Relax & Validate
Ideal for AI-predicted antibody structures (AlphaFold, RoseTTAFold, ABB3, IgFold). Upload VH/VL-only PDB files — fast AMBER14 + OBC2 energy minimisation removes prediction artefacts before geometry scoring. For full antibody-antigen complexes, use Complex Relax & Validate below.
⏱ ~2–3 min on CPU · VH/VL antibody (~3,500 atoms)
Docking Prep Pipeline
Screen for structural issues, validate antibody CDR loop geometry, then export a clean structure compatible with HADDOCK, ClusPro, ZDOCK, HDOCK, LightDock, SwarmDock, and ColabDock.
⏱ ~15–30 s on CPU
SnugDock Prep Pipeline
Screen for issues, then rename chains to L (light), H (heavy), A (antigen) and reorder for Rosetta SnugDock antibody–antigen docking.
⏱ ~10–20 s on CPU
MD Simulation Pipeline
Screen for issues, validate antibody CDR loop geometry, then run full MD prep: fix missing atoms, add hydrogens at physiological pH, solvate with TIP3P water, add neutralising ions, and energy minimise with OpenMM. Compatible with AMBER, GROMACS, NAMD.
⏱ ~5–10 min on CPU · includes CDR validation, solvation & minimisation
Complex Relax & Validate PAID
Full-accuracy energy minimisation for antibody-antigen complexes and docked structures. Uses AMBER14 + GBn2 implicit solvent (highest accuracy mode). Upload a complete complex PDB including both antibody and antigen chains.
⏱ Files with 5,000+ atoms may take 30–60 min on CPU. Add your email in session settings to be notified when results are ready.
ABB3 + Batch Merge
Predict antibody Fv structures from paired VH/VL sequences, then merge each with a shared anchor antigen PDB — ideal for complex assembly and docking prep.
⏱ ~1–2 min per sequence on CPU · up to 5 sequence pairs
Custom Pipeline Builder PAID
Build your own processing pipeline by chaining any combination of tools in any order. Each step receives the output of the previous step as its input.
Available Steps
Your Pipeline (0 steps)
Add steps from the left →
AMBER Simulation Wizard PAID
A guided workflow that prepares every attribute your structure needs before running an AMBER molecular dynamics simulation — from raw PDB to validated, parameter-ready input files.
Upload & Quality Screen
Upload your PDB. StructSure screens for gaps, missing atoms, insertion codes, altlocs, chain overlaps, and CONECT records that would cause AMBER to fail.
Automated Cleanup
Remove HETATM/solvent, resolve altlocs, strip insertion codes and OXT atoms. Produce a clean protein-only PDB suitable for LEaP.
Add Hydrogens at Target pH
Protonate the structure at your chosen pH using pdbfixer — sets histidine tautomers, protonates termini correctly.
Solvation & Ion Addition
Wrap the structure in a water box (TIP3P / OPC / SPC/E), add Na⁺/Cl⁻ ions to achieve target ionic strength and charge neutrality.
Energy Minimization
Run a short OpenMM minimization to relax clashes before handing off to AMBER — avoids instability in the first MD steps.
Download Parameter Files
Download the simulation-ready PDB with solvated, minimized coordinates. Use AMBER’s tleap locally to generate .prmtop / .inpcrd from this output.
Eastman P et al. OpenMM 7: Rapid development of high performance algorithms for molecular dynamics. PLOS Comput Biol 13(7):e1005659 (2017). DOI:10.1371/journal.pcbi.1005659
Maier JA et al. ff14SB: Improving the accuracy of protein side chain and backbone parameters from ff99SB. J Chem Theory Comput 11:3696–3713 (2015). DOI:10.1021/acs.jctc.5b00255
Import Antigen FREE
Search the RCSB PDB and AlphaFold databases for antigen structures. Download directly or send straight to Merge to build an antibody-antigen complex.