Feng Yang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Feng Yang has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Juan Liu
Wuhan University
4 shared publications - Yueyi Cai
School of Artificial Intelligence, Wuhan University, No. 299, Bayi Road, Luojia Hill Street, Wuhan, Hubei, 430072, China
2 shared publications - Jian Zhang
School of Computer Science, Wuhan University,Wuhan,China,430072
1 shared publication Jiaxing Zhang
School of Information Engineering, Hebei GEO University, Shijiazhuang, China
1 shared publicationJing Feng
Wuhan University, Wuhan, Hubei, China
1 shared publication- Jun-Feng Qu
School of Computer Engineering, Hubei University of Arts and Science, Xiangyang, Hubei, China
1 shared publication
Latest Publications
- Open Access
- 2 October 2026
- Published by Oxford University Press (OUP) in Bioinformatics
- Vol. 42 (10)
Joint enzyme–reaction retrieval and catalytic optima prediction via multimodal fusion
Motivation: Enzyme–reaction retrieval prioritizes candidate biocatalysts for experimental follow-up, where recommendations should indicate catalytic compatibility and suitable pH and temperature conditions. Existing retrieval models optimize matching scores, whereas catalytic optima predictors are typically enzyme-level regressors because condition annotations are sparse. This separation leaves ranked pairs unevaluated for condition suitability and optima predictions without reaction context. Results: We present GERO, a multimodal fusion framework integrating enzyme sequence semantics, sequence-derived pocket geometry, and molecular reaction representations through feature-gated fusion for condition-aware retrieval and catalytic optima estimation. We define the tolerance-restricted hit rate (Hit@k-TR), which requires both correct top-k retrieval and condition predictions within predefined tolerances. Across enzyme- and reaction-similarity splits, GERO improves Hit@k-TR over two-stage baselines under the primary tolerances. Benchmark examples and an iodinin biosynthesis case study illustrate candidate ranking and assay-condition estimation for experimental prioritization. Availability and implementation: Source code is available at https://github.com/ykxhs/GERO.
- 2 October 2026
- Published by Oxford University Press (OUP) in Bioinformatics
- Vol. 42 (10)
Open AccessJoint enzyme–reaction retrieval and catalytic optima prediction via multimodal fusion
Motivation: Enzyme–reaction retrieval prioritizes candidate biocatalysts for experimental follow-up, where recommendations should indicate catalytic compatibility and suitable pH and temperature conditions. Existing retrieval models optimize matching scores, whereas catalytic optima predictors are typically enzyme-level regressors because condition annotations are sparse. This separation leaves ranked pairs unevaluated for condition suitability and optima predictions without reaction context. Results: We present GERO, a multimodal fusion framework integrating enzyme sequence semantics, sequence-derived pocket geometry, and molecular reaction representations through feature-gated fusion for condition-aware retrieval and catalytic optima estimation. We define the tolerance-restricted hit rate (Hit@k-TR), which requires both correct top-k retrieval and condition predictions within predefined tolerances. Across enzyme- and reaction-similarity splits, GERO improves Hit@k-TR over two-stage baselines under the primary tolerances. Benchmark examples and an iodinin biosynthesis case study illustrate candidate ranking and assay-condition estimation for experimental prioritization. Availability and implementation: Source code is available at https://github.com/ykxhs/GERO.
- Open Access
Joint enzyme-reaction retrieval and catalytic optima prediction via multimodal fusion
Motivation: Enzyme-reaction retrieval is increasingly used to prioritize candidate biocatalysts for experimental follow-up, where useful recommendations should indicate not only whether an enzyme can catalyze a target reaction but also under which pH and temperature conditions it should be tested. Existing retrieval models optimize catalytic matching scores, whereas catalytic optima predictors are typically developed as enzyme-level regressors because public pH and temperature annotations are sparse and often available only at the enzyme or EC-associated record level. This separation leaves a practical gap: high-ranking enzyme-reaction pairs are not evaluated for condition suitability, and enzyme-level optima predictions do not use the reaction context being retrieved. Results: We present GERO, a multimodal fusion framework that uses feature-gated cross-modal fusion to integrate global enzyme sequence semantics, sequence-derived pocket geometry, and molecular reaction representations for condition-aware enzyme-reaction retrieval and catalytic optima estimation with reaction context. To evaluate this setting, we define the tolerance-restricted hit rate (Hit@k-TR), which requires both top-k retrieval of the correct candidate and condition prediction within predefined tolerances. Across enzyme- and reaction-similarity splits, GERO improves Hit@k-TR over two-stage retrieval-then-prediction baselines. Representative benchmark examples and an iodinin biosynthesis case study further illustrate GERO’s ability to provide candidate rankings together with plausible assay-condition estimates for downstream experimental prioritization. Availability and implementation: Source code is available at https://github.com/ykxhs/GERO . Contact: liujuan@whu.edu.cn Supplementary information: Supplementary data are available at XXXX online.
Open AccessJoint enzyme-reaction retrieval and catalytic optima prediction via multimodal fusion
Motivation: Enzyme-reaction retrieval is increasingly used to prioritize candidate biocatalysts for experimental follow-up, where useful recommendations should indicate not only whether an enzyme can catalyze a target reaction but also under which pH and temperature conditions it should be tested. Existing retrieval models optimize catalytic matching scores, whereas catalytic optima predictors are typically developed as enzyme-level regressors because public pH and temperature annotations are sparse and often available only at the enzyme or EC-associated record level. This separation leaves a practical gap: high-ranking enzyme-reaction pairs are not evaluated for condition suitability, and enzyme-level optima predictions do not use the reaction context being retrieved. Results: We present GERO, a multimodal fusion framework that uses feature-gated cross-modal fusion to integrate global enzyme sequence semantics, sequence-derived pocket geometry, and molecular reaction representations for condition-aware enzyme-reaction retrieval and catalytic optima estimation with reaction context. To evaluate this setting, we define the tolerance-restricted hit rate (Hit@k-TR), which requires both top-k retrieval of the correct candidate and condition prediction within predefined tolerances. Across enzyme- and reaction-similarity splits, GERO improves Hit@k-TR over two-stage retrieval-then-prediction baselines. Representative benchmark examples and an iodinin biosynthesis case study further illustrate GERO’s ability to provide candidate rankings together with plausible assay-condition estimates for downstream experimental prioritization. Availability and implementation: Source code is available at https://github.com/ykxhs/GERO . Contact: liujuan@whu.edu.cn Supplementary information: Supplementary data are available at XXXX online.
- 8 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Biology and Bioinformatics
- Vol. 22 (2) , 603-613
TRGOA: Topological-Aware Residue-Gene Ontology Attention Network for Protein Function Prediction
Protein function prediction is one of the most important biological problems in the field of bioinformatics. The functions of proteins are generally described by a series of Gene Ontology (GO) terms that have hierarchical relationships. Two factors hinder the effective prediction of protein functions using current methods: 1) they cannot well model and learn the topological semantic similarity between residues and GO terms, resulting in a huge semantic gap; 2) they predict the functions of proteins by calculating the semantic similarity between protein-level embeddings and GO terms, which does not effectively learn the protein-function relationship. To address the above issues, we propose the Topological-aware Residue-Gene Ontology Attention Network (TRGOA) for protein function prediction. Firstly, a topological-aware attention module is designed to leverage attention scores within this joint semantic space allowing for modeling the fine-grained semantic similarity between residues and GO terms, thereby narrowing the semantic gap. Secondly, a multi-head aggregator is proposed, which adeptly captures the functions relevant fine-grained semantic similarity and filters out function-irrelevant components, which effectively reveal protein-function relationships, thereby enhancing generality and robustness. Finally, TRGOA has demonstrated promising outcomes, revealing our model can understand the protein-function relationship in deep insights.
- 8 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Biology and Bioinformatics
- Vol. 22 (2) , 603-613
TRGOA: Topological-Aware Residue-Gene Ontology Attention Network for Protein Function Prediction
Protein function prediction is one of the most important biological problems in the field of bioinformatics. The functions of proteins are generally described by a series of Gene Ontology (GO) terms that have hierarchical relationships. Two factors hinder the effective prediction of protein functions using current methods: 1) they cannot well model and learn the topological semantic similarity between residues and GO terms, resulting in a huge semantic gap; 2) they predict the functions of proteins by calculating the semantic similarity between protein-level embeddings and GO terms, which does not effectively learn the protein-function relationship. To address the above issues, we propose the Topological-aware Residue-Gene Ontology Attention Network (TRGOA) for protein function prediction. Firstly, a topological-aware attention module is designed to leverage attention scores within this joint semantic space allowing for modeling the fine-grained semantic similarity between residues and GO terms, thereby narrowing the semantic gap. Secondly, a multi-head aggregator is proposed, which adeptly captures the functions relevant fine-grained semantic similarity and filters out function-irrelevant components, which effectively reveal protein-function relationships, thereby enhancing generality and robustness. Finally, TRGOA has demonstrated promising outcomes, revealing our model can understand the protein-function relationship in deep insights.
- Open Access
- 1 March 2023
- Published by Tsinghua University Press in Big Data Mining and Analytics
- Vol. 6 (1) , 1-10
FingerDTA: A Fingerprint-Embedding Framework for Drug-Target Binding Affinity Prediction
Many efforts have been exerted toward screening potential drugs for targets, and conducting wet experiments remains a laborious and time-consuming approach. Artificial intelligence methods, such as Convolutional Neural Network (CNN), are widely used to facilitate new drug discovery. Owing to the structural limitations of CNN, features extracted from this method are local patterns that lack global information. However, global information extracted from the whole sequence and local patterns extracted from the special domain can influence the drugtarget affinity. A fusion of global information and local patterns can construct neural network calculations closer to actual biological processes. This paper proposes a Fingerprint-embedding framework for Drug-Target binding Affinity prediction (FingerDTA), which uses CNN to extract local patterns and utilize fingerprints to characterize global information. These fingerprints are generated on the basis of the whole sequence of drugs or targets. Furthermore, FingerDTA achieves comparable performance on Davis and KIBA data sets. In the case study of screening potential drugs for the spike protein of the coronavirus disease 2019 (COVID-19), 7 of the top 10 drugs have been confirmed potential by literature. Ultimately, the docking experiment demonstrates that FingerDTA can find novel drug candidates for targets. All codes are available at http://lanproxy.biodwhu.cn:9099/mszjaas/FingerDTA.git.
- 1 March 2023
- Published by Tsinghua University Press in Big Data Mining and Analytics
- Vol. 6 (1) , 1-10
Open AccessFingerDTA: A Fingerprint-Embedding Framework for Drug-Target Binding Affinity Prediction
Many efforts have been exerted toward screening potential drugs for targets, and conducting wet experiments remains a laborious and time-consuming approach. Artificial intelligence methods, such as Convolutional Neural Network (CNN), are widely used to facilitate new drug discovery. Owing to the structural limitations of CNN, features extracted from this method are local patterns that lack global information. However, global information extracted from the whole sequence and local patterns extracted from the special domain can influence the drugtarget affinity. A fusion of global information and local patterns can construct neural network calculations closer to actual biological processes. This paper proposes a Fingerprint-embedding framework for Drug-Target binding Affinity prediction (FingerDTA), which uses CNN to extract local patterns and utilize fingerprints to characterize global information. These fingerprints are generated on the basis of the whole sequence of drugs or targets. Furthermore, FingerDTA achieves comparable performance on Davis and KIBA data sets. In the case study of screening potential drugs for the spike protein of the coronavirus disease 2019 (COVID-19), 7 of the top 10 drugs have been confirmed potential by literature. Ultimately, the docking experiment demonstrates that FingerDTA can find novel drug candidates for targets. All codes are available at http://lanproxy.biodwhu.cn:9099/mszjaas/FingerDTA.git.
- Open Access
- 21 October 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 194015-194026
An Angle-Based Bi-Objective Evolutionary Algorithm for Many-Objective Optimization
One of the main difficulties in solving many-objective optimization is the lack of selection pressure. For an optimization problem, its main purpose is to obtain a nondominated solution set with better convergence and diversity. In this paper, two estimation methods are proposed to convert a many-objective optimization problem into a simple bi-objective optimization problem, that is, the convergence and diversity estimation methods, so as to greatly improve the probability of certain dominance relation between solutions, and then increase the selection pressure. Based on the proposed estimation methods, a new many-objective evolutionary algorithm, termed ABOEA, is proposed. In the convergence estimation method, we use a modified ASF function to solve the performance degradation of the traditional norm distance on the irregular Pareto front. In the diversity estimation method, we innovatively propose a diversity estimation method based on the angle between solutions. Empirical experimental results demonstrate that the proposed algorithm shows its competitiveness against the state-of-art algorithms in solving many-objective optimization problems. Two estimation methods proposed in this paper can greatly improve the performance of algorithms in solving many-objective optimization problems.
- 21 October 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 194015-194026
Open AccessAn Angle-Based Bi-Objective Evolutionary Algorithm for Many-Objective Optimization
One of the main difficulties in solving many-objective optimization is the lack of selection pressure. For an optimization problem, its main purpose is to obtain a nondominated solution set with better convergence and diversity. In this paper, two estimation methods are proposed to convert a many-objective optimization problem into a simple bi-objective optimization problem, that is, the convergence and diversity estimation methods, so as to greatly improve the probability of certain dominance relation between solutions, and then increase the selection pressure. Based on the proposed estimation methods, a new many-objective evolutionary algorithm, termed ABOEA, is proposed. In the convergence estimation method, we use a modified ASF function to solve the performance degradation of the traditional norm distance on the irregular Pareto front. In the diversity estimation method, we innovatively propose a diversity estimation method based on the angle between solutions. Empirical experimental results demonstrate that the proposed algorithm shows its competitiveness against the state-of-art algorithms in solving many-objective optimization problems. Two estimation methods proposed in this paper can greatly improve the performance of algorithms in solving many-objective optimization problems.