ApexFold
Predict peptide structural redistribution across environments.
Results
Loading starting example…
Method & evidence
Estimate CD-like alpha-helical, beta-like, and unstructured fractions in four fixed environments, then compare solvent-driven shifts and structural plasticity through the qualified CPU ecosystem and production ESM API.
Scope & limitations
- Predictions cover four fixed experimental environments and should not be extrapolated to arbitrary solvent compositions.
- Outputs describe ensemble-level secondary-structure fractions; they do not resolve atomic conformations, turns, or irregular structure.
Interpreting results
- Fractions are model predictions resembling CD-derived ensemble summaries, not atomic structures or measured spectra.
- SDS denotes the paper’s 10 mmol/L membrane-mimetic micelle condition, not a homogeneous solvent titration.
- Mixed-state peptides were the hardest reported class; use predictions to prioritize experiments rather than replace CD measurements.
References
Input requirements & outputs
Input checks
- Only the 20 canonical amino acids
- Between 8 and 50 residues per peptide
- Up to 16 rows and 800 total residues in one Run
Expected results
- Four secondary-structure fraction vectors — Compare alpha-helical, beta-like, and unstructured fractions in water, methanol/water, TFE/water, and SDS micelles.
- Structural plasticity — Review generalized Jensen–Shannon distance across the four predicted compositional vectors.
- Water-to-environment alpha shifts — Inspect predicted changes in alpha-helical fraction relative to water for each nonaqueous condition.
