Toward the third generation artificial intelligence

B Zhang, J Zhu, H Su - Science China Information Sciences, 2023 - Springer
There have been two competing paradigms in artificial intelligence (AI) development ever
since its birth in 1956, ie, symbolism and connectionism (or sub-symbolism). While …

[PDF][PDF] Deep learning in protein structural modeling and design

W Gao, SP Mahajan, J Sulam, JJ Gray - Patterns, 2020 - cell.com
Deep learning is catalyzing a scientific revolution fueled by big data, accessible toolkits, and
powerful computational resources, impacting many fields, including protein structural …

Learning from protein structure with geometric vector perceptrons

B Jing, S Eismann, P Suriana… - International …, 2020 - openreview.net
Learning on 3D structures of large biomolecules is emerging as a distinct area in machine
learning, but there has yet to emerge a unifying network architecture that simultaneously …

Protein representation learning by geometric structure pretraining

Z Zhang, M Xu, A Jamasb… - arXiv preprint arXiv …, 2022 - arxiv.org
Learning effective protein representations is critical in a variety of tasks in biology such as
predicting protein function or structure. Existing approaches usually pretrain protein …

[PDF][PDF] 迈向第三代人工智能

张钹, 朱军, 苏航 - 中国科学: 信息科学, 2020 - scis.scichina.com
摘要人工智能(artificial intelligence, AI) 自1956 年诞生以来, 在60 多年的发展历史中,
一直存在两个相互竞争的范式, 即符号主义与连接主义(或称亚符号主义). 二者虽然同时起步 …

Multi-scale representation learning on proteins

VR Somnath, C Bunne… - Advances in Neural …, 2021 - proceedings.neurips.cc
Proteins are fundamental biological entities mediating key roles in cellular function and
disease. This paper introduces a multi-scale graph construction of a protein–HoloProt …

Improved protein structure refinement guided by deep learning based accuracy estimation

N Hiranuma, H Park, M Baek, I Anishchenko… - Nature …, 2021 - nature.com
We develop a deep learning framework (DeepAccNet) that estimates per-residue accuracy
and residue-residue distance signed error in protein models and uses these predictions to …

Structure-based protein design with deep learning

S Ovchinnikov, PS Huang - Current opinion in chemical biology, 2021 - Elsevier
Since the first revelation of proteins functioning as macromolecular machines through their
three dimensional structures, researchers have been intrigued by the marvelous ways the …

Shape-based generative modeling for de novo drug design

M Skalic, J Jiménez, D Sabbadin… - Journal of chemical …, 2019 - ACS Publications
In this work, we propose a machine learning approach to generate novel molecules starting
from a seed compound, its three-dimensional (3D) shape, and its pharmacophoric features …

Atom3d: Tasks on molecules in three dimensions

RJL Townshend, M Vögele, P Suriana, A Derry… - arXiv preprint arXiv …, 2020 - arxiv.org
Computational methods that operate on three-dimensional molecular structure have the
potential to solve important questions in biology and chemistry. In particular, deep neural …