ScanNet

What is ScanNet?

ScanNet (Spatio-Chemical Arrangement of Neighbors Network) is a geometric deep learning model for predicting protein binding sites directly from 3D structures. It identifies where proteins interact with other proteins, antibodies, or intrinsically disordered proteins (IDPs).

Unlike traditional methods that rely on handcrafted features or structural homology, ScanNet learns spatio-chemical patterns end-to-end from atomic coordinates. The model constructs representations by examining the spatial and chemical arrangement of neighboring atoms, enabling it to detect binding sites even on novel protein folds not seen during training.

Published in Nature Methods in 2022, ScanNet demonstrated state-of-the-art accuracy on multiple benchmarks while remaining interpretable through visualization of learned filters.

Applications

Recommended ScanNet applications:

• Protein-protein interface mapping: Identifying which residues mediate interactions between protein binding partners
• Epitope prediction: Locating antibody binding sites on antigens for vaccine design and therapeutic development
• IDP binding site detection: Finding regions where structured proteins interact with intrinsically disordered proteins
• Drug target analysis: Characterizing binding surfaces for structure-based drug design
• Structural biology: Guiding mutagenesis experiments to validate predicted interaction sites

Limitations

Known ScanNet limitations:

• Static structure assumption: ScanNet analyzes a single conformational state. Proteins with multiple binding-competent conformations or significant induced-fit changes may require ensemble analysis.
• Structure-only execution: ProteinIQ currently runs ScanNet in no-MSA mode, which avoids local HHblits and sequence database requirements. Natural proteins with rich homolog coverage may benefit from MSA-enabled ScanNet runs in a separately configured environment.
• Binding partner agnostic: ScanNet predicts general binding propensity, not partner-specific interfaces. A residue predicted as high probability may participate in any protein-protein interaction, not necessarily the one of interest.
• Resolution sensitivity: Very low-resolution structures or models with significant coordinate errors may produce unreliable predictions.

How to use ScanNet online

ProteinIQ provides a web interface for running ScanNet without command-line installation or local database setup. Upload a protein structure, select a prediction mode, and receive per-residue binding probabilities.

Pricing

The current base price is 20 credits per job. The submission quote remains authoritative; adding queries to a native list does not create separate jobs.

Inputs

Single-structure mode accepts PDB, ENT, CIF and mmCIF files, or a native PDB or UniProt identifier. Files are limited to 50 MiB each. The RCSB fetch button captures an asymmetric-unit structure; it is analyzed exactly as fetched. To let ScanNet retrieve biological assembly 1, enter the PDB identifier as text.

Native query-list mode accepts a .txt file or text with one query per line, for example:

1brs_AB
1crn_A
uploaded.cif_0-AA+1-BB

End every line, including the last line, with a newline for the native prediction CLI. Upload local structures referenced by filename in the list under Structure. Up to 200 local files can be attached. The native program determines which queries succeed; successful query outputs and native warnings are retained together.

Permissive feature mode can skip failed downloads or unprocessable structures. A missing local filename still stops the native program, so every local reference must match an attached file.

UniProt identifiers retrieve the matching AF-<accession>-F1 structure through the official AlphaFold Database API. The retrieved version and URL are recorded in the run log, and the downloaded structure is retained. Upload a saved structure when a fixed data version is required.

Settings

Hosted ScanNet supports structure-only no-MSA models. API requests with noMSA=false are rejected before execution preparation.

Results

Predictions retain every native query table, annotated structure and visualization script that ScanNet writes. Multiple query tables include a query_table display column identifying their native file. Native CSV files keep their original headers, values, precision and ordering.

The feature API returns a complete native-feature-result.pkl containing the unmodified Python return value, including residue keys and query identities. NumPy results also include separate .npy arrays; residue-dictionary arrays are bundled in feature-arrays.npz. The feature-index.json file records each array's location within the return value, shape and dtype. The displayed array index is transport metadata, not an additional scientific score. Only load pickle files from trusted jobs.

Tables, annotated structures and feature values are available as typed workflow outputs. Standard output, standard error, invocation settings and acquisition identity are downloadable with the result.

Interpreting binding probabilities

Higher probabilities indicate residues ScanNet ranks as more likely to participate in a binding interface. Use the per-residue values together with the annotated structure outputs to inspect contiguous surface patches rather than relying on a single residue in isolation. ScanNet does not define universal high/medium/low probability cutoffs, so interpret the score distribution in the context of the submitted structure.

How does ScanNet work?

ScanNet employs a hierarchical architecture that processes protein structures at both atomic and amino acid scales, learning to recognize spatio-chemical patterns associated with binding interfaces.

Spatio-chemical filters

At the core of ScanNet are trainable filters that detect specific spatial arrangements of atoms with particular chemical properties. For each atom in the structure, neighboring atoms within a local coordinate frame are extracted. These point clouds pass through linear filters that respond to arrangements such as "hydrophobic atoms surrounding a polar center" or "aromatic ring flanked by charged residues."

The filters are parameterized to be interpretable—each can be visualized to understand what molecular pattern it detects. This interpretability distinguishes ScanNet from black-box deep learning approaches.

Hierarchical representation

ScanNet builds representations at two scales:

1. Atomic level: Spatio-chemical filters process local atomic neighborhoods, producing per-atom features that capture fine structural details.
2. Amino acid level: Atomic features are aggregated within each residue and combined with amino acid attributes (type, secondary structure). A second round of spatio-chemical filtering operates on amino acid neighborhoods.

This multi-scale approach allows ScanNet to integrate information from individual atomic contacts up to residue-level surface geometry.

Evolutionary information

When MSA data is available in a configured ScanNet installation, the method can incorporate position-weight matrices derived from sequence alignments. Conservation patterns provide complementary evidence because residues conserved across evolution often participate in functionally important interfaces. The ProteinIQ online job runs the no-MSA mode, which relies on structural features.

Training and evaluation

The source paper describes the training datasets and benchmark evaluation. Performance depends on the model, structure and use of evolutionary information; the hosted no-MSA models should not be assigned a benchmark score from a different configuration.

Reference: ScanNet: An interpretable geometric deep learning model for structure-based protein binding site prediction, Tubiana, Schneidman-Duhovny and Wolfson, Nature Methods (2022).