SXSheng Xu

Sheng Xu

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Sheng Xu has a total of 103 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Structure-based discovery of small molecules that allosterically modulate RNA function

    by Zhaoming Su, Zhiling Pan, Hao Wu, Han Hu, Jian Zou, Haiyun Ma, Hongshan Ni, Liu Wang, Jiaqi Zhao, Xiaoqi Sun, et al.

    RNAs play essential roles in biological processes and have emerged as promising therapeutic targets for both oligonucleotide- and small molecule-based interventions. However, targeting RNAs with small molecules remain challenging due to our limited understandings of the chemical space associated with RNA-binding small molecules, the intricate three-dimensional RNA structures and ligand-binding pockets, and the molecular basis of small-molecule modes of action (MOAs). In this study, we discovered Hit Small molecules TArgeting RNA Structures (HitSTARS) through the following three stages. We first curated a library of 5.2 million in-stock Commercial Small molecules TArgeting RNA (CSTAR) through training a molecular representation learning model based on 3D atom positions, and explored their physicochemical properties. Next, virtual screening of CSTAR against the paradigmatic Tetrahymena ScaI L-16 ribozyme, as a proof of concept, identified two novel allosteric pockets. Functional screening of the top-scoring 125 compounds, validated by multiple docking algorithms, resulted in nine hit RNA-binders capable of inhibiting catalysis. Finally, cryo-EM structures of the best hit of each allosteric site, ZPT-005 and ZPT-084, in complexes with L-16 were determined at 2.3 Å and 3.0 Å resolutions to elucidate their modes of action: ZPT-005 stabilizes J1/2 in the undocked conformation upon binding to site 1, whereas ZPT-084 occupies site 2 to displace the catalytic core away from the splice site. These results facilitated lead optimization of ZPT-084 that improved the binding affinity and activity by 104- and 15-fold. Collectively, HitSTARS provides a comprehensive framework for streamlining RNA structure-based chemical probe discovery while offering insights into their MOAs.

  • Top Cited Papers
    Open Access

    Benchmarking all-atom biomolecular structure prediction with FoldBench

    by Sheng Xu, Qiantai Feng, Lifeng Qiao, Hao Wu, Tao Shen, Yu Cheng, Shuangjia Zheng, Siqi Sun

    Accurate prediction of biomolecular complex structures is fundamental for understanding biological processes and rational therapeutic design. Recent advances in deep learning methods, particularly all-atom structure prediction models, have significantly expanded their capabilities to include diverse biomolecular entities, such as proteins, nucleic acids, ligands, and ions. However, comprehensive benchmarks covering multiple interaction types and molecular diversity remain scarce, limiting fair and rigorous assessment of model performance and generalizability. To address this gap, we introduce FoldBench, an extensive benchmark dataset consisting of 1522 biological assemblies categorized into nine distinct prediction tasks. Our evaluations reveal critical performance dependencies, showing that ligand docking accuracy notably diminishes as ligand similarity to the training set decreases, a pattern similarly observed in protein-protein interaction modeling. Furthermore, antibody-antigen predictions remain particularly challenging, with current methods exhibiting failure rates exceeding 50%. Among evaluated models, AlphaFold 3 consistently demonstrates superior accuracy across the majority of tasks. In summary, our results highlight significant advancements yet reveal persistent limitations within the field, providing crucial insights and benchmarks to inform future model development and refinement.


  • Open Access

    FoldBench: An All-atom Benchmark for Biomolecular Structure Prediction

    by Sheng Xu, Qiantai Feng, Lifeng Qiao, Hao Wu, Tao Shen, Yu Cheng, Shuangjia Zheng, Siqi Sun

    Accurate prediction of biomolecular complex structures is fundamental for understanding biological processes and rational therapeutic design. Recent advances in deep learning methods, particularly all-atom structure prediction models, have significantly expanded their capabilities to include diverse biomolecular entities, such as proteins, nucleic acids, ligands, and ions. However, comprehensive benchmarks covering multiple interaction types and molecular diversity remain scarce, limiting fair and rigorous assessment of model performance and generalizability. To address this gap, we introduce FoldBench, an extensive benchmark dataset consisting of 1,522 biological assemblies categorized into nine distinct prediction tasks. Our evaluations reveal critical performance dependencies, showing that ligand docking accuracy notably diminishes as ligand similarity to the training set decreases, a pattern similarly observed in protein-protein interaction modeling. Furthermore, antibody-antigen predictions remain particularly challenging, with current methods exhibiting failure rates exceeding 50%. Among evaluated models, AlphaFold 3 consistently demonstrates superior accuracy across the majority of tasks. In summary, our results highlight significant advancements yet reveal persistent limitations within the field, providing crucial insights and benchmarks to inform future model development and refinement.


  • Cryo-EM reveals mechanisms of natural RNA multivalency

    by Liu Wang, Jiahao Xie, Tao Gong, Hao Wu, Yifan Tu, Xin Peng, Sitong Shang, Xinyu Jia, Haiyun Ma, Jian Zou, et al.

    Homo-oligomerization of biological macromolecules leads to functional assemblies that are critical to understanding various cellular processes. However, RNA quaternary structures have rarely been reported. Comparative genomics analysis has identified RNA families containing hundreds of sequences that adopt conserved secondary structures and likely fold into complex three-dimensional structures. In this study, we used cryo–electron microscopy (cryo-EM) to determine structures from four RNA families, including ARRPOF and OLE forming dimers and ROOL and GOLLD forming hexameric, octameric, and dodecameric nanostructures, at 2.6- to 4.6-angstrom resolutions. These homo-oligomeric assemblies reveal a plethora of structural motifs that contribute to RNA multivalency, including kissing-loop, palindromic base-pairing, A-stacking, metal ion coordination, pseudoknot, and minor-groove interactions. These results provide the molecular basis of intermolecular interactions driving RNA multivalency with potential functional relevance.


  • Top Cited Papers
    Open Access

    π-PrimeNovo: an accurate and efficient non-autoregressive deep learning model for de novo peptide sequencing

    by Xiang Zhang, Tianze Ling, Zhi Jin, Sheng Xu, Zhiqiang Gao, Boyan Sun, Zijie Qiu, Jiaqi Wei, Nanqing Dong, Guangshuai Wang, et al.

    Peptide sequencing via tandem mass spectrometry (MS/MS) is essential in proteomics. Unlike traditional database searches, deep learning excels at de novo peptide sequencing, even for peptides missing from existing databases. Current deep learning models often rely on autoregressive generation, which suffers from error accumulation and slow inference speeds. In this work, we introduce π-PrimeNovo, a non-autoregressive Transformer-based model for peptide sequencing. With our architecture design and a CUDA-enhanced decoding module for precise mass control, π-PrimeNovo achieves significantly higher accuracy and up to 89x faster inference than state-of-the-art methods, making it ideal for large-scale applications like metaproteomics. Additionally, it excels in phosphopeptide mining and detecting low-abundance post-translational modifications (PTMs), marking a substantial advance in peptide sequencing with broad potential in biological research.


  • Open Access

    RTM-align: an improved RNA alignment tool with enhanced short sequence performance via post-standardization and fragment alignment

    by Zijie Qiu, Sheng Xu, Junkang Wei, Tao Shen, Siqi Sun

    Understanding the three-dimensional structure of RNA is crucial for studying various biological processes. Accurate alignment and comparison of RNA structures are essential for illustrating RNA functionality and evolution. The existing RNA alignment tools suffer from limitations such as size-dependency of scoring functions and inadequate handling of short RNA fragments, leading to the conflicting interpretation of structural and functional relationships among RNA molecules. Hence, we introduce RTM-align, a novel RNA structural alignment tool enhanced for short RNAs. RTM-align employs the RTM-score, which integrates post-standardization to ensure size-independence and utilizes a fragment alignment strategy that improves alignment accuracy by concentrating on structural motifs and local structural elements. Benchmarking results demonstrate that RTM-align outperforms existing tools in RNA structure comparison, offering a universal scoring scale regardless of RNA length. The improvement is particularly evident in evaluating predicted structures for CASP15 RNA targets, with significant enhancements observed for the short RNA target R1117. RTM-align is expected to significantly improve the accuracy and reliability of RNA structure alignment and comparison, thereby aiding in the deeper understanding and discovery of novel RNA functions and their interactions in biological systems. RTM-align is now available at https://github.com/BEAM-Labs/RTM-align .


  • Open Access

    AcrNET: predicting anti-CRISPR with deep learning

    by Yunxiang Li, Yumeng Wei, Sheng Xu, Qingxiong Tan, Licheng Zong, Jiuming Wang, Yixuan Wang, Jiayang Chen, Liang Hong, Yu Li

    Motivation: As an important group of proteins discovered in phages, anti-CRISPR inhibits the activity of the immune system of bacteria (i.e. CRISPR-Cas), offering promise for gene editing and phage therapy. However, the prediction and discovery of anti-CRISPR are challenging due to their high variability and fast evolution. Existing biological studies rely on known CRISPR and anti-CRISPR pairs, which may not be practical considering the huge number. Computational methods struggle with prediction performance. To address these issues, we propose a novel deep neural network for anti-CRISPR analysis (AcrNET), which achieves significant performance. Results: On both the cross-fold and cross-dataset validation, our method outperforms the state-of-the-art methods. Notably, AcrNET improves the prediction performance by at least 15% regarding the F1 score for the cross-dataset test problem comparing with state-of-art Deep Learning method. Moreover, AcrNET is the first computational method to predict the detailed anti-CRISPR classes, which may help illustrate the anti-CRISPR mechanism. Taking advantage of a Transformer protein language model ESM-1b, which was pre-trained on 250 million protein sequences, AcrNET overcomes the data scarcity problem. Extensive experiments and analysis suggest that the Transformer model feature, evolutionary feature, and local structure feature complement each other, which indicates the critical properties of anti-CRISPR proteins. AlphaFold prediction, further motif analysis, and docking experiments further demonstrate that AcrNET can capture the evolutionarily conserved pattern and the interaction between anti-CRISPR and the target implicitly. Availability and implementation: Web server: https://proj.cse.cuhk.edu.hk/aihlab/AcrNET/. Training code and pre-trained model are available at.


  • Open Access

    Corrigendum: Identification of an autophagy-related 12-lncRNA signature and evaluation of NFYC-AS1 as a pro-cancer factor in lung adenocarcinoma

    by Fang Tong, Lifa Xu, Sheng Xu, Mingming Zhang

    Corrigendum: Identification of an autophagy-related 12-lncRNA signature and evaluation of NFYC-AS1 as a pro-cancer factor in lung adenocarcinoma

  • Open Access

    Distance Map Supervised Landmark Localization for MR-TRUS Registration

    by Xinrui Song, Xuanang Xu, Sheng Xu, Baris Turkbey, Bradford J. Wood, Thomas Sanford, Pingkun Yan

    In this work, we propose to explicitly use the landmarks of prostate to guide the MR-TRUS image registration. We first train a deep neural network to automatically localize a set of meaningful landmarks, and then directly generate the affine registration matrix from the location of these landmarks. For landmark localization, instead of directly training a network to predict the landmark coordinates, we propose to regress a full-resolution distance map of the landmark, which is demonstrated effective in avoiding statistical bias to unsatisfactory performance and thus improving performance. We then use the predicted landmarks to generate the affine transformation matrix, which outperforms the clinicians' manual rigid registration by a significant margin in terms of TRE.

  • Open Access

    IDa-Det: An Information Discrepancy-aware Distillation for 1-bit Detectors

    by Sheng Xu, Yanjing Li, Bohan Zeng, Teli Ma, Baochang Zhang, Xianbin Cao, Peng Gao, Jinhu Lv

    Knowledge distillation (KD) has been proven to be useful for training compact object detection models. However, we observe that KD is often effective when the teacher model and student counterpart share similar proposal information. This explains why existing KD methods are less effective for 1-bit detectors, caused by a significant information discrepancy between the real-valued teacher and the 1-bit student. This paper presents an Information Discrepancy-aware strategy (IDa-Det) to distill 1-bit detectors that can effectively eliminate information discrepancies and significantly reduce the performance gap between a 1-bit detector and its real-valued counterpart. We formulate the distillation process as a bi-level optimization formulation. At the inner level, we select the representative proposals with maximum information discrepancy. We then introduce a novel entropy distillation loss to reduce the disparity based on the selected proposals. Extensive experiments demonstrate IDa-Det's superiority over state-of-the-art 1-bit detectors and KD methods on both PASCAL VOC and COCO datasets. IDa-Det achieves a 76.9% mAP for a 1-bit Faster-RCNN with ResNet-18 backbone. Our code is open-sourced on https://github.com/SteveTsui/IDa-Det.