Zhirui Liao

Fudan University

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

Shanfeng Zhu

Fudan University

Hiroshi Mamitsuka

Kyoto University Institute for Chemical Research

Xiaodi Huang

Charles Sturt University

Christos Proukakis

National Hospital for Neurology and Neurosurgery

Caoimhe Morley

University College London

Marco Toffoli

University College London

Lianrong Pu

Shandong University

Anthony H.V. Schapira

University College London

Fritz J. Sedlazeck

Baylor Genetics

Jaehwan Kim

National Institute of Animal Science

Cooperative Institutions

Research Network (United States)

Fudan University

National Hospital for Neurology and Neurosurgery

University College London

The Royal Free Hospital

King's College London

Shanghai Center for Brain Science and Brain-Inspired Technology

Shanghai Institute for Science of Science

Nanjing University

Institute of Science and Technology

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

DeepDock: Enhancing Ligand-protein Interaction Prediction by a Combination of Ligand and Structure Information

2021 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) (2019)

Zhirui Liao Ronghui You Xiaodi Huang Xiaojun Yao Tao Huang

The prediction of precise protein-ligand binding activities can accelerate drug discovery by virtual screening-a computational technique that predicts whether a small molecule ligand is able to bind to a specific target. Thus, it is crucial to improve the performance of virtual screening. However, previous models for solving this problem are either ligand-based or structure-based. In this paper, we propose a universal deep neural network model called DeepDock that predicts protein-ligand interaction by using both ligand and structure information. Using the combination of two types of information, our model consists of embedding, convolution, max pooling, and fully-connected layers. In particular, different types of inputs are concatenated before being fed into the fully-connected layers. In the experiments, we compare our approach to the competing methods against two benchmark datasets under different settings. The experiment results have demonstrated that DeepDock can improve predictive performance by more than 4% on both DUD-E and MUV datasets in terms of AUPR.

Pooling

Benchmark (surveying)

Convolution (computer science)

Cheminformatics

Protein ligand

10.1109/bibm47256.2019.8983365

Cite

Citations (28)

An auto-resize cache structure for high-performance and low-power

ITC-CSCC :International Technical Conference on Circuits Systems, Computers and Communications (2009)

Zhirui Liao Jaehwan Kim Jong‐Wha Chong

In this paper, we design a cache scheme which can reduce the power consumption and increase the performance of cache through auto-resize of L1 cache, which is called auto-selecting cache scheme. Cache memory occupies a significant fraction of a chip’s overall power consumption. Recent researches advocate using “resizable” to adjust cache capability based on the fact that requirement in applications will reduce cache size and power consumption. Based on the fact that different programs need different sizes of instruction and data cache, an auto-selecting cache scheme is proposed. This scheme can dynamically adjust the sizes of level 1 cache according to program requirement of instruction and data cache. The proposed structure can reduce power consumption and improve cache performance. According to the SPEC2000 simulation, the average power consumption of L1 cache is reduced by 7.43% and the average of energy delay production is improved by 16.08% with auto-selecting cache structure compared traditional one.

Cache invalidation

Smart Cache

Cache pollution

Page cache

MESI protocol

Source

Cite

Citations (0)

Drug3D-DTI: Improved Drug-target Interaction Prediction by Incorporating Spatial Information of Small Molecules

Zhirui Liao Xiaodi Huang Hiroshi Mamitsuka Shanfeng Zhu

Cite

Citations (0)

Sc2Mol: a scaffold-based two-step molecule generator with variational autoencoder and transformer

Bioinformatics (2022)

Zhirui Liao Lei Xie Hiroshi Mamitsuka Shanfeng Zhu

Finding molecules with desired pharmaceutical properties is crucial in drug discovery. Generative models can be an efficient tool to find desired molecules through the distribution learned by the model to approximate given training data. Existing generative models (i) do not consider backbone structures (scaffolds), resulting in inefficiency or (ii) need prior patterns for scaffolds, causing bias. Scaffolds are reasonable to use, and it is imperative to design a generative model without any prior scaffold patterns.We propose a generative model-based molecule generator, Sc2Mol, without any prior scaffold patterns. Sc2Mol uses SMILES strings for molecules. It consists of two steps: scaffold generation and scaffold decoration, which are carried out by a variational autoencoder and a transformer, respectively. The two steps are powerful for implementing random molecule generation and scaffold optimization. Our empirical evaluation using drug-like molecule datasets confirmed the success of our model in distribution learning and molecule optimization. Also, our model could automatically learn the rules to transform coarse scaffolds into sophisticated drug candidates. These rules were consistent with those for current lead optimization.The code is available at https://github.com/zhiruiliao/Sc2Mol.Supplementary data are available at Bioinformatics online.

Autoencoder

Generative model

10.1093/bioinformatics/btac814

Cite

Citations (10)

VACmap: An Accurate Long-Read Aligner for Unraveling Complex Genomic Rearrangements

Research Square (Research Square) (2025)

Shanfeng Zhu Hongyu Ding Fritz J. Sedlazeck Christos Proukakis Caoimhe Morley

Abstract Inversions, duplications, and other critical medically challenging variations are often ignored by routine genetic analyses. This is due to the complexity of these alleles but also because of the inability to accurately align them with state-of-the-art methods. We introduce VACmap, a novel non-linear long-read mapping method designed to improve detection of these difficult genomic regions, including critical genes like LPA and GBA1, which are significant risk factors for cardiovascular disease and Parkinson’s disease, respectively.

10.21203/rs.3.rs-5088253/v1

Cite

Citations (0)

VACmap: An Accurate Long-Read Aligner for Unraveling Complex Genomic Rearrangements

bioRxiv (Cold Spring Harbor Laboratory) (2023)

Hongyu Ding Fritz J. Sedlazeck Christos Proukakis Caoimhe Morley Marco Toffoli

Inversions, duplications, and other critical medically challenging variations are often ignored by routine genetic analyses. This is due to the complexity of these alleles but also because of the inability to accurately align them with state-of-the-art methods. We introduce VACmap, a novel non-linear long-read mapping method designed to improve detection of these difficult genomic regions, including critical genes like LPA and GBA1, which are significant risk factors for cardiovascular disease and Parkinson disease, respectively.

10.1101/2023.08.03.551566

Cite

Citations (0)

Drug3D-DTI: Improved Drug-target Interaction Prediction by Incorporating Spatial Information of Small Molecules

2021 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) (2021)

Zhirui Liao Xiaodi Huang Hiroshi Mamitsuka Shanfeng Zhu

A number of machine learning (ML) approaches for drug discovery have been available that rely only on sequential (1D) and planar (2D) information without effectively using the 3D information for generating features of drugs. However, 3D information of small molecules can reflect relative position of atoms more directly, which affects molecular properties. In this work, we present a new deep learning model called Drug3D-DTI for drug-target interaction prediction. Drug3D-DTI takes advantage of molecular spatial information, i.e., atom proximity in three-dimensional (3D) structures. We comprehensively evaluated the performance of Drug3D-DTI on two datasets with two tasks of regression and classification. In particular, we compared Drug3D-DTI with several existing methods including the two cutting-edge methods for compound-protein interaction prediction. From the experimental results, Drug3D-DTI clearly outperformed other methods under all settings. Further, this performance improvement was validated by ablation experiments and a case study. The implementation of Drug3D-DTI is available at (https://github.com/zhiruiliao/Drug3D-DTI).

10.1109/bibm52615.2021.9669707

Cite

Citations (3)