Zeyu Luo

Zeyu Luo

Auto-generated profile, based on publicly available information

Zeyu Luo

College of Life Science, Chongqing Normal University , No. 37 University Town Road, high-tech District, Chongqing 401331

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
22 January 2024 in Briefings in Bioinformatics
Verified
Journal Article
Open access
PDF file
ZLZeyu Luo
Zeyu Luo
RWRui Wang
Rui Wang
YSYawen Sun
Yawen Sun
JLJunhao Liu
Junhao Liu
ZCZongqing Chen
Zongqing Chen
YZYu-Juan Zhang
Yu-Juan Zhang
As the application of large language models (LLMs) has broadened into the realm of biological predictions, leveraging their capacity for self-supervised learning to create feature representations of amino acid sequences, these models have set a new benchmark in tackling downstream challenges, such as subcellular localization. However, previous studies have primarily focused on either the structural design of models or differing strategies for fine-tuning, largely overlooking investigations into the nature of the features derived from LLMs. In this research, we propose different ESM2 representation extraction strategies, considering both the character type and position within the ESM2 input sequence. Using model dimensionality reduction, predictive analysis and interpretability techniques, we have illuminated potential associations between diverse feature types and specific subcellular localizations. Particularly, the prediction of Mitochondrion and Golgi apparatus prefer segments feature closer to the N-terminal, and phosphorylation site-based features could mirror phosphorylation properties. We also evaluate the prediction performance and interpretability robustness of Random Forest and Deep Neural Networks with varied feature inputs. This work offers novel insights into maximizing LLMs’ utility, understanding their mechanisms, and extracting biological domain knowledge. Furthermore, we have made the code, feature extraction API, and all relevant materials available at https://github.com/yujuan-zhang/feature-representation-for-LLMs.

24 May 2024 in Desalination
Verified
Journal Article
Closed access
SGShiwei Guo
Shiwei Guo
XYXinlu Yan
Xinlu Yan
ZLZeyu Luo
Zeyu Luo
JZJunhou Zhang
Junhou Zhang
CYChungang Yuan
Chungang Yuan

2 July 2025 in Briefings in Bioinformatics
Verified
Journal Article
Open access
YSYawen Sun
Yawen Sun
RWRui Wang
Rui Wang
ZLZeyu Luo
Zeyu Luo
LTLejia Tan
Lejia Tan
JLJunhao Liu
Junhao Liu
YZYu-Juan Zhang
Yu-Juan Zhang
The prediction of binary protein–protein interactions (PPIs) is essential for protein engineering, but a major challenge in deep learning-based methods is the unknown decision-making process of the model. To address this challenge, we propose the ESM2_AMP framework, which utilizes the ESM2 protein language model for extracting segment features from actual amino acid sequences and integrates the Transformer model for feature fusion in binary PPIs prediction. Further, the two distinct models, ESM2_AMPS and ESM2_AMP_CSE are developed to systematically explore the contributions of segment features and combine with special tokens features in the decision-making process. The experimental results reveal that the model relying on segment features demonstrates strong correlations between segments with high attention weights and known functional regions of amino acid sequences. This insight suggests that attention to these segments helps capture biologically relevant functional and interaction-related information. By analyzing the coverage relationship between high-attention sequence fragments and functional regions, we validated the model’s ability to capture key segment features of PPIs and revealed the critical role of functional domains in PPIs. This finding not only enhances the interpretability methods for sequence-based prediction models but also provides biological evidence supporting the important regulatory role of functional sequences in protein–protein interactions. It offers cross-disciplinary insights for algorithm optimization and experimental validation research in the field of computational biology.

26 June 2025 in Bioinformatics
Verified
Journal Article
Open access
MLMeiling Liu
Meiling Liu
SWShuangshuang Wang
Shuangshuang Wang
ZLZeyu Luo
Zeyu Luo
GWGuo-Hua Wang
Guo-Hua Wang
YZYuming Zhao
Yuming Zhao
Motivation: Protein function prediction is important for drug development and disease treatment. Recently, deep learning methods have leveraged protein sequence and structural information, achieving remarkable progress in the field of protein function prediction. However, existing methods ignore the complex multimodal interaction information between sequence and structural features. Since protein sequence and structural information reveal the functional characteristics of proteins from different perspectives, it is challenging to effectively fuse the information from these two modalities to portray protein functions more comprehensively. In addition, current methods have difficulty in effectively capturing long-range dependencies and global contextual information in protein sequences during feature extraction, thus limiting the ability of the model to recognize critical functional residues. Results: In this study, we propose a novel framework termed Multi-stage Attention-based Extraction and Fusion model for GO prediction (MAEF-GO) based on a multistage attention mechanism to predict protein functions. MAEF-GO innovatively integrates the graph convolutional network and the graph attention network to extract protein structural features. To address the issue of modeling long-range dependencies within protein sequences, we introduce a frequency-domain attention mechanism capable of extracting global contextual relationships. Additionally, a cross-attention module is implemented to facilitate interactive fusion between protein sequence and structural modalities. Experimental evaluations demonstrate that MAEF-GO achieves superior performance compared to several state-of-the-art baseline models across standard benchmarks. Furthermore, analysis of the cross-attention weight distributions demonstrates MAEF-GO’s interpretability. It can effectively identify critical functional residues of proteins. Availability and implementation: The MAEF-GO source code can be found at https://github.com/nebstudio/MAEF-GO, an archived snapshot of the code used in this study is also available via Zenodo at https://doi.org/10.5281/zenodo.15422392.

22 July 2023 in Journal of Computational Biophysics and Chemistry
Verified
Journal Article
Closed access
PDF file
MZMan Zhang
Man Zhang
CZChengxu Zhu
Chengxu Zhu
ZLZeyu Luo
Zeyu Luo
JLJunhao Liu
Junhao Liu
Muhammad Tahir KhanMuhammad Tahir Khan
Muhammad Tahir Khan
YZYu-Juan Zhang
Yu-Juan Zhang

21 November 2025 in New Microbes and New Infections
Verified
Journal Article
Open access
Muhammad Tahir KhanMuhammad Tahir Khan
Muhammad Tahir Khan
ZLZeyu Luo
Zeyu Luo
AKArwa Omar Al Khatib
Arwa Omar Al Khatib
DADalal Sulaiman Alshaya
Dalal Sulaiman Alshaya
AAAhmed Al-Qahtani
Ahmed Al-Qahtani
TNTariq Nadeem
Tariq Nadeem
Whole-genome sequencing is the most promising approach for public health surveillance and antimicrobial drug resistance. The current study aimed to analyze base mutations across Mycobacterium tuberculosis genomes to assess mutation frequency and distribution across different regions. This study aimed to identify genomic regions of Mycobacterium tuberculosis with varying mutation frequencies to inform mechanisms of drug resistance and potential drug target discovery. The study analyzed base mutations across 209 whole genome sequences, which were aligned with reference H37Rv (NC_000962.3), using the PhyResSE pipeline. Based on the frequency of mutations, the regions have been classified into three main locations. High frequency mutation areas: around 2300 kb to 2400 kb, around 4100 kb to 4200 kb, around 1600 kb to 1700 kb, and 3700 kb to 3800 kb. These locations showed dense clusters linked to katG and inhA (Isoniazid resistance), Ethambutol resistance (embB), rifampicin resistance (compensatory role rpoA). Moderate frequency mutations were observed around 2000 kb to 2100 kb, around 1300 kb to 1400 kb, around 3800 kb to 3900 kb, and around 3400 kb to 3500 kb. These regions show mostly involved rrs mutations (amikacin, kanamycin, and capreomycin resistance), and rpoA has a compensatory role in rifampicin resistance. Regions exhibiting minimal mutation activity have been observed around 100 kb–300 kb, around 500 kb, and 2700 kb to 2800 kb. The most common nucleotide substitution was G to A (G→A) (16 %), followed by A to G (A→G) and T to C (T→C) (15 %). These findings collectively highlight novel genomic regions of stability and hypervariability, offering insights for refining WGS-based surveillance, resistance prediction, and future drug target prioritization.