Protein Mutation Folding Stability (Physics-based ΔΔG)

Protein Mutation Folding Stability (Physics-based ΔΔG)

Predict how a single amino acid mutation changes a protein's folding stability.

Get started How to run it

Protein Mutation Folding Stability (Physics-based ΔΔG) calculates the effect of a single amino acid substitution on protein folding stability. The result is reported as the change in folding free energy, ΔΔG_fold, in kcal/mol.

Under the reported sign convention, a negative ΔΔG_fold indicates stabilisation of the folded state, while a positive value indicates destabilisation.

Supported mutation targets include the twenty natural amino acids and fourteen unnatural amino acids: Aib, Sar, dA, dF, dW, dY, hPhe, Hyp, mePhe, meS, meT, Nle, Orn, and Phe_4F.

When to use it

This calculation is applicable to the assessment of point mutations in protein engineering, including the stabilisation of enzymes and antibodies and the evaluation of unnatural amino acids at defined sites. This version accepts single-chain protein structures. Multichain structures for which the mutation assignment is ambiguous are rejected.

Inputs

Input	Required	What it is
protein_pdb	yes	The input protein structure in PDB format. This version accepts a single chain.
mutations	no	One substitution in the form CHAIN:RESID:TARGET, for example A:78:V. Unnatural amino acid targets may be written bare or in square brackets, for example A:78:Aib or A:78:[Aib]. The job computes ΔΔG for a single substitution. The mutations and mutant_chain_helm inputs are mutually exclusive.
mutant_chain_helm	no	The complete mutated chain in HELM2 notation. Mutation sites are assigned by position-wise alignment with the input PDB chain. The mutant_chain_helm and mutations inputs are mutually exclusive.
mutant_chain_id	no, default A	The PDB chain corresponding to the sequence supplied in mutant_chain_helm.
unfolded_flank_size	no, default 3	The number of residues retained on each side of the mutation site in the unfolded reference peptide. The default of 3 defines a 7-residue window. A value of 0 defines a single capped residue.
unfolded_engine	no, default feflow	The engine used for the unfolded leg. The accepted values are feflow, which is recommended, and rfe_legacy, which retains the previous calculation path as a fallback.
unfolded_relax_ns	no, default 0.005 ns	The reference-peptide relaxation duration before the free energy calculation, in nanoseconds.
unfolded_relax_implicit_solvent	no, default true	Controls the solvent representation during reference-peptide relaxation. A value of true selects implicit solvent. A value of false selects explicit water, as specified by the published quantitative protocol, with a corresponding increase in runtime.
unfolded_relax_restrained_nvt_ns	no, default 0.0 ns	The duration of an additional backbone-restrained relaxation phase, in nanoseconds. A value of 0.0 skips this phase.
unfolded_relax_unrestrained_nvt_ns	no, default 0.0 ns	The duration of an additional unrestrained relaxation phase following the restrained phase, in nanoseconds. A value of 0.0 skips this phase.
mode_preset	no, default smoke	Applies a predefined set of parameter values. smoke applies no overrides and uses the supplied values, which correspond to the defaults in this table unless modified. aldeghi_2019_quantitative applies the published Aldeghi 2019 protocol by increasing the relaxation duration, selecting explicit water for the reference peptide, and reducing the flanks to a single capped residue.
random_seed	no, default 42	The fixed random seed used to obtain reproducible results across repeated submissions with identical inputs.
equil_length_ns	no, default 5.0 ns	The equilibration duration per endpoint, in nanoseconds. The minimum value is 0.005, which is intended for pipeline verification and is unsuitable for scientific interpretation.
n_neq_switches_per_direction	no, default 50	The number of switching calculations performed in each direction. Increasing this value improves the statistical estimate and increases computational cost.
protocol_repeats	no, default 3, minimum 1	The number of independent repeats included in the uncertainty estimate. Increasing this value can reduce statistical uncertainty and increases computational cost.

Longer simulations and larger numbers of switches and repeats increase runtime and credit consumption while improving statistical reliability.

How to run it

Submissions can be created through Azulene Studio, the Python SDK, or the CLI. The Get started page contains installation, authentication, and initial submission instructions.

In Azulene Studio

Select Protein Mutation Folding Stability (Physics-based ΔΔG) from the tools list. In the Inputs and Parameters step, upload the protein PDB, enter the mutation specification, such as A:6:F, and configure the simulation settings or select a mode_preset. Complete the submission through Review and Submit.

From the Python SDK

Submit the input structure and mutation through the Python SDK.

from azulene import jobs

result = jobs.submit(
    job_type="protein_mutation_ddg_fold",
    input_data={
        "protein_pdb": "/path/to/your/protein.pdb",
        "mutations": "A:6:F",
        "protocol_repeats": 3,
    },
)

From the CLI

Pass the inputs as a JSON string. The protein PDB path is uploaded automatically.

azulene jobs submit --job-type protein_mutation_ddg_fold \
  --input-data '{"protein_pdb": "/path/to/your/protein.pdb", "mutations": "A:6:F", "protocol_repeats": 3}'

Reading the result

The primary output is ddG_fold, the mutation-induced change in folding free energy in kcal/mol. Its uncertainty estimate is reported as sigma_total. A negative ddG_fold indicates increased protein stability, while a positive value indicates decreased protein stability.

The folded and unfolded contributions are also reported separately. dG_folded is the free energy change for the mutation in the folded protein, with uncertainty sigma_folded and unit dG_folded_unit. dG_unfolded is the corresponding free energy change for the unfolded reference peptide, with uncertainty sigma_unfolded and unit dG_unfolded_unit. The reported ddG_fold equals dG_folded minus dG_unfolded.

Notes

The defaults execute a full-length protocol comprising 5 ns of equilibration per endpoint, 50 switches per direction, and 3 repeats. They are not a cheap trial run. The smoke preset name does not reduce these values. A lower-cost initial pass requires explicit reductions to equil_length_ns, n_neq_switches_per_direction, and protocol_repeats; the resulting output should be treated solely as a pipeline verification result.

Use aldeghi_2019_quantitative for comparison with published quantitative data. Runtime scales with simulation duration, switch count, and repeat count. This version requires a single-chain input protein.