Protein Features and Structure: parameters and notes

See the tutorial for usage and examples.

Each function includes its purpose, API, inputs, outputs, example code and result description. Prepare any required credentials, local files and optional dependencies; remote results vary with database releases and query time.

These functions call existing models or prediction services. Weights, SDKs, credentials and service terms remain provider requirements; DNAKit did not retrain these models.

Label biological categories of protein SAE features

dnakit.databases.esm_describe_sae_feature(*, feature_id, sae_model=None, model=None, n_proteins=None, top_residues_per_protein=None, use_cache=None, verbose=True)

Parameter	Type	Required / default	Description and constraints
feature_id	integer	Required	SAE feature index in [0, 16383]. Use the values returned by ESM_get_sae_features.active_features[].feature_id.
sae_model	string	None	SAE checkpoint to label. Cache is keyed on this — different SAEs produce different labels for the same id.
model	string	None	ESMC backbone.
n_proteins	integer	None	Number of panel proteins to run SAE on. Default 10 (the full curated panel). Lower for cheaper exploration at the cost of confidence.
top_residues_per_protein	integer	None	For each protein, take the top-K residues where the target feature activates most strongly, then check their UniProt annotations. Default 3.
use_cache	boolean	None	If true (default) and a cached label exists at ~/.cache/tooluniverse/sae_labels/.../feature_{id}.json, return it instead of rerunning the panel. Set false to force a recompute.
verbose	bool	True	Progress flag.

Configure before running: ESM_API_KEY.

Aggregate protein SAE features over a residue interval

dnakit.databases.esm_get_region_sae_features(*, sequence, start_position, end_position, top_k_features=None, model=None, sae_model=None, verbose=True)

Parameter	Type	Required / default	Description and constraints
sequence	string	Required	Protein amino acid sequence in single-letter code. Up to ~2700 AA.
start_position	integer	Required	1-indexed inclusive start of the region of interest.
end_position	integer	Required	1-indexed inclusive end of the region of interest. Must be >= start_position and within sequence length. Region length cap: 500.
top_k_features	integer	None	Number of top features (by total |activation| over region) to return.
model	string	None	ESMC base model
sae_model	string	None	SAE codebook identifier
verbose	bool	True	Progress flag.

Configure before running: ESM_API_KEY.

Extract per-residue protein SAE activations

dnakit.databases.esm_get_sae_features(*, sequence, model=None, sae_model=None, position=None, window=None, top_k_per_residue=None, verbose=True)

Parameter	Type	Required / default	Description and constraints
sequence	string	Required	Protein amino acid sequence in single-letter code (e.g. 'MEEPQSDPSVEPPLSQETFSDLWKLLPENN'). Standard 20 amino acids; avoid gaps or non-standard characters. Practical length cap ~2,700 AA.
model	string	None	ESMC backbone model. Currently SAE is only released for esmc-6b-2024-12 (default).
sae_model	string	None	SAE checkpoint name. Default is the layer-60 SAE matching esmc-6b backbone.
position	integer | null	None	Optional 1-indexed residue position. If set, only activations within +/- window residues of this position are returned (use for variant analysis to keep response compact). If null, all residues are returned (large response for long sequences).
window	integer	None	Residue window radius around position (only used when position is set). Default 8 = +/- 8 residues = 17-residue window centered on the variant.
top_k_per_residue	integer	None	Cap features returned per residue, sorted by absolute activation. Default 64 = full sparsity (no cap). Set lower (e.g. 16) to reduce response size when no position filter is used.
verbose	bool	True	Progress flag.

Configure before running: ESM_API_KEY.

Score protein sequence log-probabilities with ESMC

dnakit.databases.esm_score_sequence(*, sequence, model=None, verbose=True)

Parameter	Type	Required / default	Description and constraints
sequence	string	Required	Protein amino acid sequence in single-letter code to score (e.g. 'MKTAYIAKQRQISFVKSHFSRQ').
model	string	None	ESMC model to use for scoring.
verbose	bool	True	Progress flag.

Configure before running: ESM_API_KEY.

Predict protein structures with AlphaFold2

dnakit.databases.nvidia_nim_alphafold2(*, sequence, algorithm=None, databases=None, e_value=None, iterations=None, relax_prediction=None, skip_template_search=None, verbose=True)

Parameter	Type	Required / default	Description and constraints
sequence	string	Required	Amino acid sequence to predict structure for (single letter codes)
algorithm	string	None	MSA search algorithm. mmseqs2 is faster, jackhmmer is more sensitive
databases	array	None	Sequence databases for MSA search: uniref90, mgnify, small_bfd {"items": {"type": "string"}, "default": ["small_bfd"]}
e_value	number	None	E-value threshold for MSA search
iterations	integer	None	Number of search iterations
relax_prediction	boolean	None	Whether to perform structure relaxation
skip_template_search	boolean	None	Skip template-based prediction
verbose	bool	True	Progress flag.

Configure before running: NVIDIA_API_KEY.

Predict protein structures from sequence with ESMFold

dnakit.databases.nvidia_nim_esmfold(*, sequence, verbose=True)

Parameter	Type	Required / default	Description and constraints
sequence	string	Required	Amino acid sequence (max 1024 chars, valid: ARNDCQEGHILKMFPSTWYVXBOU)
verbose	bool	True	Progress flag.

Configure before running: NVIDIA_API_KEY.