← Research overview · Full article
Inspect individual files or download the complete archive. File formats are plain CSV and PDB, with a text README and license.
Download all supplementary data (ZIP) · README · CC BY 4.0
The experimental files report measured outcomes. Structural metrics describe computational predictions. The four starting reference complexes preserve archived scaffold residue names, which may differ from the actual sequence. Use the sequence CSV for sequence identity.
| File | Contents |
|---|---|
| experimental_binding_results.csv40 data rows | All 40 submitted binder-core sequences, binding outcomes and tag positions. |
| experimental_assay_conditions.csv40 data rows | Exported target concentrations, replicate counts and assay classifications. |
| supplementary_bli_lineage.csv20 data rows | Exact lineage for the 20 submitted optimized sequences; missing prediction matches remain missing. |
| representative_sequence_pairs.csv4 data rows | Four parent/optimized sequence pairs used in Figures 2 and 3. These sequences are authoritative. |
| representative_reference_structures.csv4 data rows | Reference chain assignments and target sequences. Scaffold residue names can differ from the parent sequence. |
| representative_prediction_metrics.csv56 data rows | Fresh-prediction metrics for the four representative pairs across seven models. |
| supplementary_site_retention.csv2,520 data rows | Residue-contact retention data for the protein-target comparisons. |
| supplementary_nonprotein_targets.csv120 data rows | All 120 DNA/small-molecule redesign outputs and computational metrics. |
| nonprotein_representative_metrics.csv2 data rows | Paired sequences and metrics for the nonprotein examples in Figure S5. |
These four structures define the starting geometry for the representative examples. They are not relabeled as predictions of the paired parent sequences.
Article and supplementary data: CC BY 4.0. Research software has separate license terms.