Haozhou Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Haozhou Li has a total of 36 co-authors. Here are the top co-authors by number of shared publications:
- Qinke Peng
Systems Engineering Institute, Xian Jiaotong University, Xian, China
12 shared publications - Ruimeng Li
Systems Engineering Institute, Xian Jiaotong University, Xian, China
6 shared publications - Zeyuan Zeng
Systems Engineering Institute, Xian Jiaotong University, Xian, China
5 shared publications - Xu Mou
Systems Engineering Institute, Xian Jiaotong University, Xian, China
4 shared publications Wentong Sun
Systems Engineering Institute, Xian Jiaotong University, Xian, China
3 shared publications- Ying Wang
School of Automation Science and Engineering, Xian Jiaotong University, Xian, Shannxi, China
3 shared publications
Latest Publications
- 28 March 2026
- Published by Elsevier in Applied Soft Computing
- Vol. 196
LumiCRS: Asymmetric contrastive prototype learning for long-tail conversational recommender systems
- 28 March 2026
- Published by Elsevier in Applied Soft Computing
- Vol. 196
LumiCRS: Asymmetric contrastive prototype learning for long-tail conversational recommender systems
- Open Access
- 28 February 2026
- Published by Springer Nature in BMC Bioinformatics
- Vol. 27 (1)
Multimodal learning on heterogeneous subgraphs and LLMs representation for MHC-peptide binding affinity prediction
Accurate prediction of MHC-peptide binding affinity remains a challenge for immunotherapeutic development. Existing methods struggle to jointly model functional semantics of polymorphic residues, evolutionary conservation constraints, and structural dynamic. We propose the Contrast learning-based Multi-feature Heterogeneous Subgraph model (CMHS) with sequence and structural representation. For sequence representation, we introduce LoRA fine-tuning to obtain the MHC-exclusive sequence representation from ESM2, then jointly BLOSUM50 to capture long-range functional dependencies and evolutionarily conserved residues. For structural representation, we use the biophysics-guided heterogeneous graph network. Constructing an MHC-peptide graph with a novel trainable Gaussian noise layer guided by crystallographic B-factors to dynamically simulate electron density uncertainty, coupled with a three-stage message-passing framework with subgraph aggregation, subgraph extraction and heterogeneous. Finally, to align sequence and graph representation spaces, we use contrastive learning to obtain a more comprehensive representation and to enhance the ability of model prediction. Evaluations on 16 HLA allele benchmarks show average SRCC improvements of 8.7 $$\%$$ , with improvements of average AUC of 7.6 $$\%$$ . This work establishes a new paradigm for predicting hypervariable immune interactions. The corresponding code can be founded in github.- 28 February 2026
- Published by Springer Nature in BMC Bioinformatics
- Vol. 27 (1)
Open AccessMultimodal learning on heterogeneous subgraphs and LLMs representation for MHC-peptide binding affinity prediction
Accurate prediction of MHC-peptide binding affinity remains a challenge for immunotherapeutic development. Existing methods struggle to jointly model functional semantics of polymorphic residues, evolutionary conservation constraints, and structural dynamic. We propose the Contrast learning-based Multi-feature Heterogeneous Subgraph model (CMHS) with sequence and structural representation. For sequence representation, we introduce LoRA fine-tuning to obtain the MHC-exclusive sequence representation from ESM2, then jointly BLOSUM50 to capture long-range functional dependencies and evolutionarily conserved residues. For structural representation, we use the biophysics-guided heterogeneous graph network. Constructing an MHC-peptide graph with a novel trainable Gaussian noise layer guided by crystallographic B-factors to dynamically simulate electron density uncertainty, coupled with a three-stage message-passing framework with subgraph aggregation, subgraph extraction and heterogeneous. Finally, to align sequence and graph representation spaces, we use contrastive learning to obtain a more comprehensive representation and to enhance the ability of model prediction. Evaluations on 16 HLA allele benchmarks show average SRCC improvements of 8.7 $$\%$$ , with improvements of average AUC of 7.6 $$\%$$ . This work establishes a new paradigm for predicting hypervariable immune interactions. The corresponding code can be founded in github. - 20 January 2026
- Published by Elsevier in Information Processing & Management
- Vol. 63 (4)
SEHLP: A summary-enhanced large language model for financial report sentiment analysis via hybrid LoRA and dynamic prefix tuning
- 20 January 2026
- Published by Elsevier in Information Processing & Management
- Vol. 63 (4)
SEHLP: A summary-enhanced large language model for financial report sentiment analysis via hybrid LoRA and dynamic prefix tuning
- 25 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Biology and Bioinformatics
- Vol. 22 (6) , 3175-3187
Dynamic Graph Attention Meets Pretrained Language Models: Adaptive K-Mer Decomposition for LncRNA-Protein Interaction Prediction
Protein-RNA complexes, particularly those involving RNA-binding proteins and long non-coding RNAs (lncRNA), are commonly found to influence gene expression and mediate fundamental cellular processes. Despite significant advances in representations for these biological sequences, sequence decomposition based on k-mer generally results in fix-length substrings, failing to detect the information of variable-length biological functional regions. In this paper, we develop a concept of expressiveness for k-mer decompositions as a theoretical underpinning for traversing all k-mer decompositions. Based on this concept, we propose an advanced approach, BERTDGA-LPI, to detect the information of variable-length biological functional regions utilizing dynamic graph attention and to capture the influence of RNA and protein context leveraging pretrained language models. The experimental results demonstrate the outperformance of BERTDGA-LPI over state-of-the-art methods across two homo sapiens datasets, one plant species dataset, and two species-unspecific datasets. Furthermore, BERTDGA-LPI is validated as effective in predicting unknown RNA-protein interactions (RPI) with 100% prediction accuracy in six independent validation sets from different species. This study lays a theoretical underpinning for traversing all k-mer decompositions and innovatively offers a broadly applicable and efficient tool for LPI prediction and RPI prediction based only on sequences.
- 25 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Biology and Bioinformatics
- Vol. 22 (6) , 3175-3187
Dynamic Graph Attention Meets Pretrained Language Models: Adaptive K-Mer Decomposition for LncRNA-Protein Interaction Prediction
Protein-RNA complexes, particularly those involving RNA-binding proteins and long non-coding RNAs (lncRNA), are commonly found to influence gene expression and mediate fundamental cellular processes. Despite significant advances in representations for these biological sequences, sequence decomposition based on k-mer generally results in fix-length substrings, failing to detect the information of variable-length biological functional regions. In this paper, we develop a concept of expressiveness for k-mer decompositions as a theoretical underpinning for traversing all k-mer decompositions. Based on this concept, we propose an advanced approach, BERTDGA-LPI, to detect the information of variable-length biological functional regions utilizing dynamic graph attention and to capture the influence of RNA and protein context leveraging pretrained language models. The experimental results demonstrate the outperformance of BERTDGA-LPI over state-of-the-art methods across two homo sapiens datasets, one plant species dataset, and two species-unspecific datasets. Furthermore, BERTDGA-LPI is validated as effective in predicting unknown RNA-protein interactions (RPI) with 100% prediction accuracy in six independent validation sets from different species. This study lays a theoretical underpinning for traversing all k-mer decompositions and innovatively offers a broadly applicable and efficient tool for LPI prediction and RPI prediction based only on sequences.
- 30 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Biology and Bioinformatics
- Vol. 22 (5) , 2143-2151
Compound Interaction Presentation Learning for MHC-Peptide Binding Affinity Prediction
The interaction between peptides and Major Histocompatibility Complex Class I (MHC-I) molecules plays a critical role in adaptive immune recognition. Although computational prediction algorithms have advanced over traditional experimental methods, challenges still remain. There is a scarcity of standardized datasets that provide comprehensive profiles of MHC-peptide structure. The polymorphism of MHC molecules introduces diverse binding patterns, complicating the characterization of specific amino acid interaction pairs. To address these issues, we introduce GSM, a novel deep learning model that combines a Graph Attention Neural Network with a Self-Attention Convolutional Neural Network to predict MHC-peptide binding affinities. By integrating self-attention mechanisms to capture global peptide-MHC interactions and graph-based modeling to represent local amino acid pairwise interactions, GSM provides a comprehensive understanding of binding mode. Compared to existing algorithms, GSM shows superior performance and greater stability across diverse allele datasets, as demonstrated on the benchmarks. Furthermore, by leveraging real 3D structural data and attention visualization, GSM is capacity of selecting interaction sites, offering valuable insights for vaccine design and advancing immunological research.
- 30 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Biology and Bioinformatics
- Vol. 22 (5) , 2143-2151
Compound Interaction Presentation Learning for MHC-Peptide Binding Affinity Prediction
The interaction between peptides and Major Histocompatibility Complex Class I (MHC-I) molecules plays a critical role in adaptive immune recognition. Although computational prediction algorithms have advanced over traditional experimental methods, challenges still remain. There is a scarcity of standardized datasets that provide comprehensive profiles of MHC-peptide structure. The polymorphism of MHC molecules introduces diverse binding patterns, complicating the characterization of specific amino acid interaction pairs. To address these issues, we introduce GSM, a novel deep learning model that combines a Graph Attention Neural Network with a Self-Attention Convolutional Neural Network to predict MHC-peptide binding affinities. By integrating self-attention mechanisms to capture global peptide-MHC interactions and graph-based modeling to represent local amino acid pairwise interactions, GSM provides a comprehensive understanding of binding mode. Compared to existing algorithms, GSM shows superior performance and greater stability across diverse allele datasets, as demonstrated on the benchmarks. Furthermore, by leveraging real 3D structural data and attention visualization, GSM is capacity of selecting interaction sites, offering valuable insights for vaccine design and advancing immunological research.
- 26 June 2025
- Published by Elsevier in Information Processing & Management
- Vol. 62 (6)
CSTE: A Context-enhanced Speaker-aware Triple Encoding model for intelligent triage and diagnosis in medical dialogue
- 26 June 2025
- Published by Elsevier in Information Processing & Management
- Vol. 62 (6)
CSTE: A Context-enhanced Speaker-aware Triple Encoding model for intelligent triage and diagnosis in medical dialogue
- 9 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. 29 (9) , 6940-6952
A Finetuning Deep Learning Framework for Pan-Species Promoters Identification With Pseudo Time Series Analysis on Time and Frequency Space
Promoter identification and classification play crucial roles in unraveling gene mechanisms. Promoters are characterized by specific motifs, such as the TATA-box for eukaryotes and the Pribnow box for prokaryotes, which are known as elements. These constitute the core components, intimately tied to promoter function. However, the heterogeneity of promoters across different species poses a significant challenge to improving identification models. In our study, we introduce ProTriCNN, a deep learning method designed for promoter identification. Based on promoters representation, ProTriCNN treats promoters as pseudo-time series, utilizing this approach to capture the intricate heterogeneity of promoter elements. Furthermore, we introduce TransPro, a ProTriCNN-based Fine-tuning framework to improve identification performance across different species. To better align source species and target species, the TransPro utilizes elements and species evolutionary trees to represent the locality difference between source and target species across various levels and time-frequency space, respectively. Compared to state-of-the-art methods, ProTriCNN demonstrates superior performance across all species, achieving an average accuracy improvement of 2.1% and a 20% enhancement in the Matthews coefficient. TransPro further attains accuracy improvement of the highest 8% and a 25% enhancement in the Matthews coefficient compared to ProTriCNN. The source code and the associated datasets are freely available at https://github.com/Limomo33/promoter.
- 9 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. 29 (9) , 6940-6952
A Finetuning Deep Learning Framework for Pan-Species Promoters Identification With Pseudo Time Series Analysis on Time and Frequency Space
Promoter identification and classification play crucial roles in unraveling gene mechanisms. Promoters are characterized by specific motifs, such as the TATA-box for eukaryotes and the Pribnow box for prokaryotes, which are known as elements. These constitute the core components, intimately tied to promoter function. However, the heterogeneity of promoters across different species poses a significant challenge to improving identification models. In our study, we introduce ProTriCNN, a deep learning method designed for promoter identification. Based on promoters representation, ProTriCNN treats promoters as pseudo-time series, utilizing this approach to capture the intricate heterogeneity of promoter elements. Furthermore, we introduce TransPro, a ProTriCNN-based Fine-tuning framework to improve identification performance across different species. To better align source species and target species, the TransPro utilizes elements and species evolutionary trees to represent the locality difference between source and target species across various levels and time-frequency space, respectively. Compared to state-of-the-art methods, ProTriCNN demonstrates superior performance across all species, achieving an average accuracy improvement of 2.1% and a 20% enhancement in the Matthews coefficient. TransPro further attains accuracy improvement of the highest 8% and a 25% enhancement in the Matthews coefficient compared to ProTriCNN. The source code and the associated datasets are freely available at https://github.com/Limomo33/promoter.
- Open Access
SubstationAI: Multimodal Large Model-Based Approaches for Analyzing Substation Equipment Faults
The reliability of substation equipment is crucial to the stability of power systems, but traditional fault analysis methods heavily rely on manual expertise, limiting their effectiveness in handling complex and large-scale data. This paper proposes a substation equipment fault analysis method based on a multimodal large language model (MLLM). We developed a database containing 40,000 entries, including images, defect labels, and analysis reports, and used an image-to-video generation model for data augmentation. Detailed fault analysis reports were generated using GPT-4. Based on this database, we developed SubstationAI, the first model dedicated to substation fault analysis, and designed a fault diagnosis knowledge base along with knowledge enhancement methods. Experimental results show that SubstationAI significantly outperforms existing models, such as GPT-4, across various evaluation metrics, demonstrating higher accuracy and practicality in fault cause analysis, repair suggestions, and preventive measures, providing a more advanced solution for substation equipment fault analysis.
Open AccessSubstationAI: Multimodal Large Model-Based Approaches for Analyzing Substation Equipment Faults
The reliability of substation equipment is crucial to the stability of power systems, but traditional fault analysis methods heavily rely on manual expertise, limiting their effectiveness in handling complex and large-scale data. This paper proposes a substation equipment fault analysis method based on a multimodal large language model (MLLM). We developed a database containing 40,000 entries, including images, defect labels, and analysis reports, and used an image-to-video generation model for data augmentation. Detailed fault analysis reports were generated using GPT-4. Based on this database, we developed SubstationAI, the first model dedicated to substation fault analysis, and designed a fault diagnosis knowledge base along with knowledge enhancement methods. Experimental results show that SubstationAI significantly outperforms existing models, such as GPT-4, across various evaluation metrics, demonstrating higher accuracy and practicality in fault cause analysis, repair suggestions, and preventive measures, providing a more advanced solution for substation equipment fault analysis.
- 14 April 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 12712-12716
SADE: A Speaker-Aware Dual Encoding Model Based on Diagbert for Medical Triage and Pre-Diagnosis
Medical triage and diagnosis systems are crucial in alleviating the severe burden on healthcare systems, providing great assistance and convenience to patients with limited medical knowledge. Most current studies, however, struggle to effectively learn speaker-aware information in patient-doctor dialogues. Besides, the professional terms and disease-related expressions make it difficult for traditional methods to understand medical knowledge. To solve these challenges, we propose a novel Speaker-Aware Dual Encoding (SADE) model, which employs two encoders of identical structure to separately represent patient consultations and doctor diagnoses. To grasp medical knowledge, we introduce DiagBERT, pre-trained on numerous domain-specific texts to represent input utterances. We then integrate BiLSTM and Dendrite Network into each encoder to capture sequential and interactional semantics. Moreover, we have built two large-scale Triage and Diagnosis datasets to overcome the scarcity of public medical corpora. Extensive experimental results on these datasets indicate that SADE outperforms other baseline models.
- 14 April 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 12712-12716
SADE: A Speaker-Aware Dual Encoding Model Based on Diagbert for Medical Triage and Pre-Diagnosis
Medical triage and diagnosis systems are crucial in alleviating the severe burden on healthcare systems, providing great assistance and convenience to patients with limited medical knowledge. Most current studies, however, struggle to effectively learn speaker-aware information in patient-doctor dialogues. Besides, the professional terms and disease-related expressions make it difficult for traditional methods to understand medical knowledge. To solve these challenges, we propose a novel Speaker-Aware Dual Encoding (SADE) model, which employs two encoders of identical structure to separately represent patient consultations and doctor diagnoses. To grasp medical knowledge, we introduce DiagBERT, pre-trained on numerous domain-specific texts to represent input utterances. We then integrate BiLSTM and Dendrite Network into each encoder to capture sequential and interactional semantics. Moreover, we have built two large-scale Triage and Diagnosis datasets to overcome the scarcity of public medical corpora. Extensive experimental results on these datasets indicate that SADE outperforms other baseline models.
- 14 February 2024
- Published by Springer Nature in Neural Computing and Applications
- Vol. 36 (12) , 6701-6719
Multi-document influence on readers: augmenting social emotion prediction by learning document interactions
- 14 February 2024
- Published by Springer Nature in Neural Computing and Applications
- Vol. 36 (12) , 6701-6719
Multi-document influence on readers: augmenting social emotion prediction by learning document interactions