Many processes in biology and drug discovery involve various 3D interactions
between different molecules, such as protein and protein, protein and small
molecule, etc. Designing a generalist model to learn universal molecular
interactions is valuable yet challenging, given that different molecules are
usually represented in different granularity. In this paper, we first propose
to universally represent a 3D molecule as a geometric graph of sets, in
contrast to conventional single-level representations. Upon the proposed
unified representation, we then propose a Generalist Equivariant Transformer
(GET) to effectively capture both sparse block-level and dense atom-level
interactions. To be specific, GET consists of a bilevel attention module, a
feed-forward module and a layer normalization module, where, notably, each
module is E(3) equivariant to meet the symmetry of 3D world. Extensive
experiments on the prediction of protein-protein affinity, ligand binding
affinity, and ligand efficacy prediction verify the effectiveness of our
proposed method against existing methods, and reveal its potential to learn
transferable knowledge across different domains and different tasks.

该论文提出一种通用的分子相互作用模型，通过将三维分子表示为几何图形集的方法，使得能够捕获稀疏块级和密集原子级交互，通过进行广泛的实验证实了该方法对不同领域和不同任务进行可转移的学习具有潜力。

通用等变转换器及其在 3D 分子相互作用学习中的应用

Generalist Equivariant Transformer Towards 3D Molecular Interaction  Learning

Effectively predicting molecular interactions has the potential to accelerate
molecular dynamics by multiple orders of magnitude and thus revolutionize
chemical simulations. Graph neural networks (GNNs) have recently shown great
successes for this task, overtaking classical methods based on fixed molecular
kernels. However, they still appear very limited from a theoretical
perspective, since regular GNNs cannot distinguish certain types of graphs. In
this work we close this gap between theory and practice. We show that GNNs with
directed edge embeddings and two-hop message passing are indeed universal
approximators for predictions that are invariant to translation, and
equivariant to permutation and rotation. We then leverage these insights and
multiple structural improvements to propose the geometric message passing
neural network (GemNet). We demonstrate the benefits of the proposed changes in
multiple ablation studies. GemNet outperforms previous models on the COLL,
MD17, and OC20 datasets by 34%, 41%, and 20%, respectively, and performs
especially well on the most challenging molecules. Our implementation is
available online.

本研究介绍了一种新型的 Geometric Message Passing Neural Network（GemNet）结构，采用了定向边嵌入和双跳消息传递，展现出图神经网络在预测分子动力学方面的强大效果和广阔前景。

GemNet: 面向分子的通用定向图神经网络

GemNet: Universal Directional Graph Neural Networks for Molecules

Molecular interaction networks are powerful resources for the discovery. They
are increasingly used with machine learning methods to predict biologically
meaningful interactions. While deep learning on graphs has dramatically
advanced the prediction prowess, current graph neural network (GNN) methods are
optimized for prediction on the basis of direct similarity between interacting
nodes. In biological networks, however, similarity between nodes that do not
directly interact has proved incredibly useful in the last decade across a
variety of interaction networks. Here, we present SkipGNN, a graph neural
network approach for the prediction of molecular interactions. SkipGNN predicts
molecular interactions by not only aggregating information from direct
interactions but also from second-order interactions, which we call skip
similarity. In contrast to existing GNNs, SkipGNN receives neural messages from
two-hop neighbors as well as immediate neighbors in the interaction network and
non-linearly transforms the messages to obtain useful information for
prediction. To inject skip similarity into a GNN, we construct a modified
version of the original network, called the skip graph. We then develop an
iterative fusion scheme that optimizes a GNN using both the skip graph and the
original graph. Experiments on four interaction networks, including drug-drug,
drug-target, protein-protein, and gene-disease interactions, show that SkipGNN
achieves superior and robust performance, outperforming existing methods by up
to 28.8\% of area under the precision recall curve (PR-AUC). Furthermore, we
show that unlike popular GNNs, SkipGNN learns biologically meaningful
embeddings and performs especially well on noisy, incomplete interaction
networks.

SkipGNN 是一种用于预测分子相互作用的图神经网络。它不仅聚合来自直接相互作用的信息，还利用了跳跃相似性，将二次跳的邻居的信息根据非线性转换融入到预测中，从而得到比现存方法更优异的结果。