K-Fold
An openly released model developed in Korea that predicts complex structures and interface confidence.
API/CLI tool ID: kfold
The input fields depend on the value of inputFormat.
Understanding inputs and outputs
Apo refers to the structure of a component protein apart from its binding partner. In stage, apo selects this preparation step, complex selects complex prediction, and all performs both. A computationally prepared apo structure is not an experimentally verified unbound structure. Colons in sequence separate chains within one complex.
ranking_score orders candidates within this model. pLDDT assesses local structure, pTM the overall arrangement, and ipTM interfaces between components. pde is a distance error in ångström, not a probability. has_clash flags atoms overlapping too closely; the absence of clashes does not establish actual binding or function.
Input settings
| Setting | Format | Required | Default | Description and constraints |
|---|---|---|---|---|
inputFormat | task | Optional | "sequence" | Input format: sequence separates chains with ':', query is a native K-Fold query file — Allowed values sequence · query |
sequence | sequence | Conditionally required | — | Separate components (chains) with ':' — Characters ACDEFGHIKLMNPQRSTVWYX:, Applies to sequence |
queryFile | file | Conditionally required | — | K-Fold query file (YAML/JSON). Specify apo structure paths inside the query. — Extensions yaml · yml · json, Applies to query |
seed | number | Optional | — | Random seed for reproducibility (blank: random) — Minimum 0 |
stage | dropdown | Optional | "all" | all = apo preparation + full complex prediction — Allowed values all · apo · complex |
kernel | dropdown | Optional | "auto" | Backend for AtlasFold apo preparation and complex prediction — Allowed values auto · triton · cuequiv · torch |
Output metrics
| Column | Display name | Range | Unit | Description |
|---|---|---|---|---|
ranking_score | Ranking score | — | — | K-Fold's own ranking score (combined across seeds and samples) |
plddt | pLDDT | 0~100 | — | — |
ptm | pTM | 0~1 | — | — |
iptm | ipTM | 0~1 | — | Inter-component interface confidence |
pde | pDE | — | Å | Predicted distance error. This metric is not normalized to 0–1. |
has_clash | Clashes | — | — | — |
See Reading results for metric definitions and interpretation. A range shows possible values; whether higher or lower is better depends on the metric.
Execution availability
Open an execution plan in your project to see available tool versions and costs. Review and approve the inputs, settings, and terms before running.
License
Code, model weights, and accompanying data have their own terms. The table summarizes terms from official sources; check your execution plan for the configuration you will use. License texts are available in Third-party notices.
| Component | License | Commercial use | Source | Notes |
|---|---|---|---|---|
| Code | Apache-2.0 | Allowed, subject to notice requirements | License text | K-Fold code is distributed under the Apache 2.0 license. |
| Code | MIT | Allowed, subject to notice requirements | License text | AtlasFold code used for sequence representation is licensed under MIT. |
| Code | MIT | Allowed, subject to notice requirements | License text | TriProRep code for structural representations uses the MIT license. |
| Code | MIT | Allowed, subject to notice requirements | License text | bio-prior is an MIT-licensed training dependency and is not included in the current inference configuration. |
| Model weights | Apache-2.0 | Allowed, subject to notice requirements | Weight distribution metadata | The distributor’s model card lists the K-Fold weights under Apache 2.0. |
| Model weights | Apache-2.0 | Allowed, subject to notice requirements | Weight distribution metadata | The asset repository has the same terms |
| Model weights | MIT | Allowed, subject to notice requirements | Weight distribution metadata | This entry covers AtlasLM weights distributed under MIT. Other distributions may have different terms. |
| Code | MPL-2.0 | Conditional; see notes | License text | gemmi is licensed under MPL-2.0. Distributing an image to customers requires providing the source for the covered files. |
atlasfold, TriProRep, and the training-only bioprior use MIT terms. gemmi uses MPL-2.0; redistribution in a container requires providing the corresponding covered source files.
Checked on 2026-10-22.