Skip to search formSkip to main contentSkip to account menu

Neural Bellman-Ford Networks: A General Graph Neural Network Framework for Link Prediction

The Neural Bellman-Ford Network (NBFNet) is proposed, a general graph neural network framework that solves the path formulation with learned operators in the generalized BellMan-Ford algorithm.
View PDF on arXiv (opens in a new tab)

Protein Representation Learning by Geometric Structure Pretraining

Experimental results show that the proposed pretraining methods outperform or are on par with the state-of-the-art sequence-based methods, while using much less data.
View PDF on arXiv (opens in a new tab)

TorchDrug: A Powerful and Flexible Machine Learning Platform for Drug Discovery

TorchDrug is a powerful and flexible machine learning platform for drug discovery built on top of PyTorch, which benchmarks a variety of important tasks in drug discovery, including molecular property prediction, pretrained molecular representations, de novo molecular design and optimization, retrosynthsis prediction, and biomedical knowledge graph reasoning.
View PDF on arXiv (opens in a new tab)

Neural-Symbolic Models for Logical Queries on Knowledge Graphs

Experiments show that GNN-QE significantly improves over previous state-of-the-art models in answering FOL queries, and can predict the number of answers without explicit supervision, and provide visualizations for intermediate variables.
View PDF on arXiv (opens in a new tab)

Learning interpretable cellular and gene signature embeddings from single-cell transcriptomic data

A new model to address challenges in scalability, model interpretability, and confounders of computational single-cell RNA-seq analyses is shown, by learning meaningful embeddings from the data that simultaneously refine gene signatures and cell functions in diverse conditions.
View on Springer (opens in a new tab)

PEER: A Comprehensive and Multi-Task Benchmark for Protein Sequence Understanding

This paper proposes a benchmark called PEER, a comprehensive and multi-task benchmark for protein understanding tasks, and evaluates different types of sequence-based methods for each task including traditional feature engineering approaches, different sequence encoding methods as well as large-scale pre-trained protein language models.
View PDF on arXiv (opens in a new tab)

Power-Law Graphs Have Minimal Scaling of Kemeny Constant for Random Walks

A theoretically guaranteed estimation algorithm is presented, which approximates the Kemeny constant for a graph in nearly linear time with respect to the number of edges, and it is shown that this approximation algorithm is both efficient and accurate.
View on ACM (opens in a new tab)

A Roadmap for Big Model

This paper discusses not only the BM technologies themselves but also the prerequisites for BM training and applications with BMs, dividing the BM review into four parts: Resource, Models, Key Technologies and Application.
View PDF on arXiv (opens in a new tab)

Structured Multi-task Learning for Molecular Property Prediction

A method called SGNN-EBM is proposed to systematically investigate the structured task modeling from two perspec-tives, which can be e-ciently trained through noise-contrastive estimation (NCE) approach.
View PDF on arXiv (opens in a new tab)

Protein Structure Representation Learning by Geometric Pretraining

Experimental results show that the proposed pretraining methods outperform or are on par with the state-of-the-art sequence-based methods using much less data.
By clicking accept or continuing to use the site, you agree to the terms outlined in our Privacy Policy (opens in a new tab), Terms of Service (opens in a new tab), and Dataset License (opens in a new tab)