- Rating
- 3
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
The paper presents an expressivity analysis of content–geometry interactions, showing that separable additive aggregation cannot distinguish changes in their pairing and characterising bilinear interactions through outer products. Motivated by this analysis, the authors introduce Fold-Conv, which uses GK-Net to combine learned basis matrices according to relative geometry and sequence separation. They build Hyper-Fold backbones and a dedicated pocket-detection head, reporting strong results on enzyme-function prediction, fold classification, and pocket detection with better computational efficiency.
Critical strengths
-
An efficient convolution motivated by an expressivity analysis. The authors analyse an expressivity ladder from additive message passing to scalar-, channel-, and matrix-gated interactions, highlighting how these operators couple residue content with relative geometry. This analysis motivates Fold-Conv, which allows geometry to control both the weighting and mixing of residue feature channels. GK-Net generates coefficients conditioned on relative geometry and sequence separation to combine a small set of learned basis matrices, providing a practical implementation of matrix gating without generating a full transformation matrix independently for every edge.
-
Strong empirical results across multiple tasks. Hyper-Fold achieves strong results on enzyme-function prediction, fold classification, and pocket detection. Hyper-Fold-Pocket also outperforms the evaluated baselines on two external pocket benchmarks, with fewer parameters and lower reported inference latency than UniSite-3D, without pretrained protein-language-model embeddings.
Critical weaknesses
-
The theoretical expressivity analysis does not fully support the claimed expressivity hierarchy or the motivation for matrix gating. The authors use the expressivity analysis to motivate Fold-Conv, but the expressive-ceiling result applies to a restricted bilinear function class, and its extension to general multilayer geometric GNNs is not established. Moreover, distinguishing binding swaps does not by itself justify matrix gating, since simpler scalar and channel gates can already distinguish such configurations. Matrix gating allows richer geometry-dependent channel mixing, but the analysis does not clearly establish which additional interactions are important for protein modelling. This leaves a gap between the theoretical motivation and the specific operator proposed.
-
The experiments do not sufficiently isolate the additional benefit of matrix gating. The operator comparisons use a shared architecture and training recipe, which is a useful control. However, channel gating accounts for much of the improvement, while matrix gating introduces additional parameters and the reported differences lack uncertainty estimates. It therefore remains unclear whether the incremental gains reliably reflect the richer interaction mechanism rather than increased capacity or variation across runs. Furthermore, the comparison with UniSite-3D changes both the backbone and pocket-detection head, making it difficult to attribute the overall improvement specifically to Fold-Conv. The authors need more controlled comparisons to establish the contributions of the operator and the detector design separately.
Questions for authors
-
“Hypergraph modeling” in the title may be misleading. The authors explicitly state that hyperedges only organize sequence and spatial-contact neighborhoods, while Fold-Conv aggregates pairwise messages without hypergraph propagation.
-
Scope of the expressivity analysis. Could the authors clarify the scope of Theorem 1 and its relationship to existing edge-conditioned message-passing formulations? In particular, how does the claim in Appendix F.3 that stacking preserves the bilinear class account for composition and nonlinear transformations?
-
Controlled comparisons. In Table 3, matrix gating improves EC Fmax from 0.777 to 0.789 and pocket [email protected] from 0.432 to 0.467 over channel gating, while the reported classification-model parameter count increases from 3.6M to 6.6M. Could the authors provide parameter-matched comparisons with repeated runs and uncertainty estimates to establish whether these gains reflect the interaction mechanism rather than increased capacity? For pocket detection, could alternative backbones be evaluated under the same detection head to separate Fold-Conv’s contribution from the detector changes?
-
Robustness to rotations. Could the authors clarify the orientation term in Equation 6 and verify rotational invariance of both the backbone and complete pocket model? With defined as a frame-axis vector, the stated sum of element-wise products is not generally rotation-invariant. The pocket head’s Cartesian Fourier positional encoding also does not itself guarantee rotational invariance.
-
Additional capability of matrix gating. Since scalar and channel gating can already distinguish binding swaps, could the authors clarify the contrary statement about scalar gating in Section 4.3? What additional interactions does matrix gating capture, and could the authors illustrate their importance with a concrete protein-modelling example?
7 comments
Response to Q2: Scope of Theorem 1
Theorem 1 characterizes single-layer bilinear messages with a fixed geometry feature map: any bilinear map in $(x_j,\phi(\delta_{ij})) $ can be written as a linear readout of their outer product, and vice versa. This provides a common basis for comparing additive, scalar, channel, and matrix-gated interactions and motivates Fold-Conv's shared rank- basis implementation.
For multilayer networks, composition and nonlinearities generate new functional forms. For example, two successive interactions can produce , which lies outside the original bilinear class in . We therefore explicitly restrict the theorem to a single-layer characterization rather than a general expressivity ceiling for multilayer networks. We have added this scope clarification to the appendix and revised the abstract, Introduction, Figure 1, contributions, and Conclusion consistently.
We have also added direct comparisons with CDConv and other edge-conditioned message-passing methods in the Fold-Conv section and Related Work, highlighting our shared-basis parameterization, geometry-dependent cross-channel mixing, and unified sequence/contact-hyperedge organization.
Response to Q3: Controlled Comparisons and Backbone/Head Controls
We have added three independent runs to Table 3 under a shared backbone, head, neighborhood, and training recipe:
| Metric | Channel | Matrix |
|---|---|---|
| EC Fmax@50 | 0.771 ± 0.003 | 0.788 ± 0.001 |
| EC AUPR | 0.868 ± 0.005 | 0.871 ± 0.002 |
| Pocket [email protected] | 0.604 ± 0.002 | 0.615 ± 0.005 |
| Pocket [email protected] | 0.440 ± 0.002 | 0.470 ± 0.004 |
Figure 10 further expands the base width, increasing Channel to 5.5M parameters for EC and 8.9M for Pocket, comparable in scale to Matrix at 6.6M and 10.1M. At these similar scales, Matrix achieves EC Fmax@50/AUPR of 0.777/0.873, versus 0.771/0.869 for Channel, and Pocket [email protected]/[email protected] of 0.618/0.473, versus 0.606/0.447. Thus, the Matrix advantage persists after capacity expansion.
The new Table 4 separately evaluates backbone and detection-head contributions:
| Backbone | DETR-style head [email protected]/[email protected] | Hyper-Fold-Pocket head [email protected]/[email protected] |
|---|---|---|
| GearNet | 0.420/0.206 | 0.434/0.224 |
| Hyper-Fold | 0.603/0.454 | 0.610/0.473 |
Hyper-Fold outperforms GearNet under both detection heads, and our head improves performance with both backbones. These controls show that the backbone and detector each contribute to the final improvement.
Response to Q1: Hypergraph Modeling
Hyper-Fold adopts a hypergraph modeling perspective to unify sequence relations and spatial-contact relations in proteins. Sequence neighborhoods and local spatial contacts are represented by hyperedges, and Fold-Conv models content–geometry interactions within these hyperedge-organized neighborhoods.
Thus, “Hypergraph Modeling” refers to the unified organization and modeling of multiple types of residue relations, rather than to one particular standard hypergraph convolution. Pairwise message computation is our specific implementation on this hypergraph structure. We have added this clarification to the Method and Related Work to make the connection between hyperedge-based relation organization and Fold-Conv message computation explicit.