Most complex structure projects fail at construct design, not data collection.
You express multiple chains, get aggregation or substoichiometric assembly, and optimize blindly for months. Standard AlphaFold-Multimer gives structure but no insight into interface reliability, flexibility, or how to design working constructs.
Now available: Complex prediction and analysis on Orbion - from sequences to construct strategy in one workflow.
Platform Capabilities:
-
AlphaFold-Multimer + dual interface confidence (chain-to-chain + geometry)
-
Mechanistic roles (receptor/ligand, scaffold/regulator/connector)
-
Flexibility mapping (truncation candidates, hinges, multi-state residues)
-
Functional predictions (PTMs, binding sites, membrane topology)
-
Integrated co-expression protocols (stoichiometry, mutations, additives)
Works for: Antibody-antigen, enzyme complexes, membrane assemblies, kinase-substrate, protein-peptide.
Example: GLP-1R–GLP-1 Complex
GPCR-peptide (1:1 hetero-oligomer). Clinically relevant (semaglutide mechanism), experimentally validated (PDB: 6X18, 5VAI), notoriously difficult (membrane protein + transient binding + DPP4 degradation + glycosylation interference).
Structure + Interface Predictions:
-
Dual confidence (chain 18.4Å, geometry 3.1Å) explains why binding interface is reliable despite conformational dynamics
-
4 flexibility zones (N-term, ICL2/3, C-term) ✅ match truncations in solved structures
Functional Sites:
-
4 disulfide bridges (matches PDB ✅)
-
N-glycosylation (N63, N82) (expression-critical ✅)
-
Phosphorylation cluster (S431-444) in desensitization region
-
Cholesterol binding sites
-
GLP-1 C-term amidation (required for activity ✅)
Each structural prediction chains into experimental decisions:
-
N-glycosylation (N63, N82) → HEK293S GnTI⁻ at 31-32°C for glycan control
-
Cholesterol binding sites → LMNG/CHS detergent + POPC nanodisc composition
-
DPP4 cleavage site → Sitagliptin (10 μM) + A8G mutation for dual protection
-
Conformational dynamics (PAE 18.4Å) → Crosslinking/SAXS validation recommendation
-
Additional parameters: DNA ratio 1:2 + Sodium butyrate (3-5 mM, 16h) + ELISA checkpoint
Result: Complete construct strategy with interconnected parameters that normally require multiple failed expression rounds. <30 minutes from sequences.
