Article
Biochemistry & Molecular Biology
Kevin McDonnell, Enda Howley, Florence Abram
Summary: Proteomics is a technique used to study system-wide protein expression, which has wide ranging applications and impacts every area of biology. De novo peptide sequencing, a popular method, is improving with the integration of machine learning. This research evaluates two algorithms for de novo peptide sequencing and explores the characteristics of tandem mass spectra. The study highlights the challenges of missing cleavage sites and noise, and provides recommendations for algorithm improvements.
COMPUTATIONAL AND STRUCTURAL BIOTECHNOLOGY JOURNAL
(2022)
Article
Biochemistry & Molecular Biology
Kevin McDonnell, Enda Howley, Florence Abram
Summary: This research compared the performance of two state-of-the-art de novo peptide sequencing algorithms, Novor and DeepNovo, with a focus on their handling of missing fragmentation cleavage sites and noise. The study found that DeepNovo performed better overall than Novor, but Novor recalled more correct amino acids when 6 or more cleavage sites were missing.
COMPUTATIONAL AND STRUCTURAL BIOTECHNOLOGY JOURNAL
(2022)
Article
Chemistry, Analytical
Cuitong He, Catherine C. L. Wong
Summary: Alternative splicing of human genes allows the production of diverse proteoforms, which play crucial roles in normal and disease physiology. However, the discovery of low-abundance proteoforms is limited by current detection methods. This study presents the development of a novel algorithm, CNovo, which outperforms existing algorithms. Furthermore, a semi-de novo sequencing algorithm, SpliceNovo, specifically designed for identifying novel junction peptides, demonstrates higher accuracy than other algorithms. Our results significantly enhance the identification of novel proteoforms through de novo sequencing.
ANALYTICAL CHEMISTRY
(2023)
Article
Computer Science, Artificial Intelligence
Rui Qiao, Ngoc Hieu Tran, Lei Xin, Xin Chen, Ming Li, Baozhen Shan, Ali Ghodsi
Summary: The research team presents a neural network-based method that can handle sequencing data of any resolution while improving the accuracy of predicted sequences. Increasing resolution in mass spectrometry provides better data for sequencing, but also raises the challenge of computational complexity in data analysis.
NATURE MACHINE INTELLIGENCE
(2021)
Article
Biochemical Research Methods
Joon-Yong Lee, Hugh D. Mitchell, Meagan C. Burnet, Ruonan Wu, Sarah C. Jenson, Eric D. Merkley, Ernesto S. Nakayasu, Carrie D. Nicora, Janet K. Jansson, Kristin E. Burnum-Johnson, Samuel H. Payne
Summary: Metaproteomics is increasingly used for high-throughput characterization of proteins in complex environments, and the creation of a sample-specific protein sequence database is crucial for accurate analysis. This study presents a de novo peptide sequencing approach to identify sample composition directly from metaproteomic data, using a deep learning model trained on diverse bacteria. The results demonstrate the potential of this method as an alternative and complementary approach to construct sample-specific protein databases, particularly in the absence of matched metagenomes.
JOURNAL OF PROTEOME RESEARCH
(2022)
Article
Biochemistry & Molecular Biology
Kevin McDonnell, Enda Howley, Florence Abram
Summary: Proteins are crucial for living cells and proteomics, the study of their expression, has diverse applications. Peptide identification is typically done by matching mass spectra to a protein database, but de novo methods can also be used. This study critically analyzes the use of artificial data for training and evaluating de novo peptide identification algorithms.
COMPUTATIONAL AND STRUCTURAL BIOTECHNOLOGY JOURNAL
(2023)
Article
Biochemical Research Methods
Denis Beslic, Georg Tscheuschner, Bernhard Y. Renard, Michael G. Weller, Thilo Muth
Summary: Monoclonal antibodies are important proteins used in research, therapeutics, and diagnostics. Mass spectrometry-based de novo protein sequencing is a valuable method to obtain amino acid sequences without prior knowledge. Evaluating different sequencing algorithms is crucial for improving accuracy and coverage.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Biochemical Research Methods
Xingyu Liao, Min Li, Junwei Luo, You Zou, Fang-Xiang Wu, Yi-Pan, Feng Luo, Jianxin Wang
Summary: This study introduces a new framework, EPGA-SC, for de novo assembly of single-cell sequencing data, overcoming challenges such as sequencing errors, biases, and repetitive regions. By classifying reads, using high precision paired-end reads from other assemblers, and developing novel algorithms for error removal and contig extension, EPGA-SC outperforms most current tools in terms of MAX contig, N50, NG50, NA50, and NGA50.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2021)
Article
Biochemistry & Molecular Biology
Ema Svetlicic, Lucija Doncevic, Luka Ozdanovac, Andrea Janes, Tomislav Tustonic, Andrija Stajduhar, Antun Lovro Brkic, Marina Ceprnja, Mario Cindric
Summary: This study proposes a mass spectrometry-based method for rapid and accurate identification of microbial pathogens, overcoming the limitations of traditional methods. The method utilizes a novel instrument for pathogen identification without the need for a lengthy culture step, and employs a more comprehensive database for species identification.
Article
Chemistry, Multidisciplinary
Ruitao Wu, Xiang Zhang, Runtao Wang, Haipeng Wang
Summary: This paper introduces a novel peptide sequencing method, Denovo-GCN, based on graph convolutional neural networks and convolutional neural networks. It can directly infer the peptide sequence from a tandem mass spectrum. By constructing an undirected graph, CNN is used to extract the features of the nodes, and the correlation between the nodes is represented by an adjacency matrix. Experiments show that Denovo-GCN outperforms DeepNovo with a relative improvement of 13.7-25.5% in terms of peptide-level recall.
APPLIED SCIENCES-BASEL
(2023)
Review
Chemistry, Medicinal
Eugene Lin, Chieh-Hsin Lin, Hsien-Yuan Lane
Summary: This article reviews the latest developments in de novo peptide and protein design using generative adversarial network (GAN) algorithms in the interdisciplinary fields of generative chemistry, machine learning, deep learning, and computer-aided drug design and discovery. It also discusses the challenges and emerging problems for future research.
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2022)
Article
Biology
Swarnalatha Yanamadala, Sivakumar Shanthirappan, Sidhika Kannan, Narendran Chiterasu, Kumaran Subramanian, Lamya Ahmed Al-Keridis, Tarun Kumar Upadhyay, Nawaf Alshammari, Mohd Saeed, Guru Prasad Srinivasan, Rohini Karunakaran
Summary: This study investigates the anticancer properties of antimicrobial peptides (AMPs) isolated from Sphaeranthus amaranthoides, a traditional medicinal plant. One specific peptide molecule showed promising anticancer properties and can potentially be used for early cancer therapy after further analysis. This research is significant for advancing cancer treatment.
Article
Biochemical Research Methods
Mei Zhao, Zhenqi Yuan, Longtao Wu, Shenghu Zhou, Yu Deng
Summary: By constructing and characterizing a mutant library of Trc promoters, a synthetic promoter library was established with a wide range of intensities. Using this library, machine learning models were built and optimized, with the XgBoost model exhibiting optimal performance in predicting the strength of artificially designed promoter sequences. This work provides a powerful platform for predictably tuning promoters to achieve optimal transcriptional strength.
ACS SYNTHETIC BIOLOGY
(2022)
Article
Genetics & Heredity
Ying Li, Jianing Zhao, Zhaoqian Liu, Cankun Wang, Lizheng Wei, Siyu Han, Wei Du
Summary: In this paper, we propose a novel computational model MEL-MP for predicting moonlighting proteins (MPs) by utilizing specific classifiers for different feature types, resulting in superior prediction performance. Through experiments, it is shown that MEL-MP outperforms the existing machine learning model MPFit, demonstrating its effectiveness in predicting MPs.
FRONTIERS IN GENETICS
(2021)
Article
Chemistry, Analytical
Chao Yang, Jian-Hui Liu, Wei-Jie Zhang, Yi-Chu Shan, Zhong-Peng Dai, Li-Hua Zhang, Yu-Kui Zhang
Summary: The newly developed de novo sequencing method based on PIPTL can significantly improve the sequencing accuracy, with 95.51% and 93.60% sequencing accuracy applied to tryptic digested bovine serum albumin and Herceptin, respectively. This method shows promise for applications in the characterization of monoclonal antibody sequences.
CHINESE JOURNAL OF ANALYTICAL CHEMISTRY
(2021)
Article
Chemistry, Analytical
Ching Tarn, Wen-Feng Zeng
Summary: This study adopts few-shot learning method to enhance the prediction accuracy of deep learning spectrum prediction, validated on multiple datasets, showing significant improvement in prediction accuracy within seconds.
ANALYTICAL CHEMISTRY
(2021)
Article
Medicine, Research & Experimental
Yuwan Zhao, Zhuo Li, Huancheng Tang, Shanhong Lin, Wenfeng Zeng, Dongcai Ye, Xin Zeng, Qiuming Luo, Jianwei Li, Zhixian Ao, Jierong Mo, Lixin Chen, Yiqiu Yang, Yunsheng Huang, Jianjun Liu
Summary: This study investigated the synthesis of a near-infrared light-sensitive NO prodrug and its effects on prostate cancer cells. The results showed that the drug effectively inhibited cell proliferation and promoted apoptosis in a concentration-dependent manner. Furthermore, in vivo experiments demonstrated the anti-cancer effects of the drug, with increased NO concentration in tumors after near-infrared light irradiation.
BIOMEDICINE & PHARMACOTHERAPY
(2021)
Article
Biochemical Research Methods
Zhen-Lin Chen, Peng-Zhi Mao, Wen-Feng Zeng, Hao Chi, Si-Min He
Summary: pDeepXL is a deep learning tool for predicting MS/MS spectra of cross-linked peptide pairs. Trained using transfer learning, it accurately predicts spectra of both noncleavable and cleavable cross-linked peptide pairs, and shows improved robustness through online fine-tuning. Integration of pDeepXL into a database search engine increases the identification of cross-link spectra by 18% on average.
JOURNAL OF PROTEOME RESEARCH
(2021)
Article
Biochemistry & Molecular Biology
Matthias Mann, Chanchal Kumar, Wen-Feng Zeng, Maximilian T. Strauss
Summary: The rapid growth of biomedical data generation and computational capabilities has led to advancements in utilizing machine learning and deep learning in proteomics for predictive modeling and biomarker discovery. These technologies are essential for improving analytical workflows and integrating multi-omics data, while also raising concerns about model transparency, explainability, and data privacy when deploying MS-based biomarkers in clinical settings.
Article
Biochemical Research Methods
Wen-Jing Zhou, Zhuo-Hong Wei, Si-Min He, Hao Chi
Summary: In this study, we developed a more comprehensive validation method pValid 2, which successfully overcame the limitations of previous validation methods by introducing a new feature, leading to improved accuracy and efficiency in identifications.
JOURNAL OF PROTEOMICS
(2022)
Article
Biochemical Research Methods
Wen-Feng Zeng, Wei-Qian Cao, Ming-Qi Liu, Si-Min He, Peng-Yuan Yang
Summary: The study introduces a new glycan-first glycopeptide search engine, pGlyco3, which can comprehensively analyze intact N- and O-glycopeptides, including glycopeptides with modified saccharide units, in a fast and accurate manner.
Article
Computer Science, Interdisciplinary Applications
Wen-Feng Zeng, Qing-Xin Chu
Summary: An antenna decoupling method based on modal control is proposed in this paper, which excites a pair of decoupling modes simultaneously to achieve decoupling. The effectiveness of this method is validated through the analysis and design of a head-to-head antenna pair. Additionally, an eight-element MIMO antenna is designed, fabricated, and measured to demonstrate the good performance of the proposed method.
INTERNATIONAL JOURNAL OF RF AND MICROWAVE COMPUTER-AIDED ENGINEERING
(2022)
Article
Biochemistry & Molecular Biology
Isabell Bludau, Sander Willems, Wen-Feng Zeng, Maximilian T. Strauss, Fynn M. Hansen, Maria C. Tanzer, Ozge Karayel, Brenda A. Schulman, Matthias Mann
Summary: The recent revolution in computational protein structure prediction has provided new insights into the study of the entire proteome. In this study, the researchers analyze posttranslational modifications (PTMs) of proteins to determine their structural context and investigate their potential regulatory sites. The analysis reveals global patterns of PTM occurrence and spatial coregulation of different types of PTMs.
Article
Engineering, Electrical & Electronic
Yu-Zhong Liang, Fu-Chang Chen, Wen-Feng Zeng, Qing-Xin Chu
Summary: This communication investigates the method of mode cancellation for designing a two-port dielectric resonator antenna (DRA) for in-band full-duplex (IBFD) applications. The antenna structure is simple, consisting only of a single DRA element, a pair of feeding lines, and a pair of metallic probes. By utilizing different modes, the mutual coupling between the exciting port and the passive port can be suppressed to a very low level without the need for an extra decoupling structure. A prototype is fabricated and measured to verify the design, with the results demonstrating broad bandwidth and high isolation throughout the working band.
IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION
(2023)
Article
Multidisciplinary Sciences
Wen-Feng Zeng, Xie-Xuan Zhou, Sander Willems, Constantin Ammar, Maria Wahle, Isabell Bludau, Eugenia Voytik, Maximillian T. Strauss, Matthias Mann
Summary: Machine learning and deep learning are becoming increasingly important in MS-based proteomics. AlphaPeptDeep is a modular Python framework built on PyTorch that can learn and predict peptide properties. It features a model shop that allows non-specialists to create models easily. AlphaPeptDeep can also predict sequence-based properties and performs well in predicting retention time, collisional cross sections, and fragment intensities.
NATURE COMMUNICATIONS
(2022)
Article
Multidisciplinary Sciences
Siyuan Kong, Pengyun Gong, Wen-Feng Zeng, Biyun Jiang, Xinhang Hou, Yang Zhang, Huanhuan Zhao, Mingqi Liu, Guoquan Yan, Xinwen Zhou, Xihua Qiao, Mengxi Wu, Pengyuan Yang, Chao Liu, Weiqian Cao
Summary: pGlycoQuant is a generic tool for quantitative analysis of intact glycopeptides using both primary and tandem mass spectrometry. It employs a deep learning model and a Match In Run algorithm to improve glycopeptide matching and expand the quantitative function of various search engines. Its application in N-glycoproteomic study demonstrates its potential in exploring site-specific glycosylation and its role in biological processes.
NATURE COMMUNICATIONS
(2022)
Article
Cardiac & Cardiovascular Systems
Nan Cai, Cunren Li, Xianfang Gu, Wenfeng Zeng, Jiawei Zhong, Jingfeng Liu, Guopeng Zeng, Junxing Zhu, Haifeng Hong
Summary: The study found that there is a relationship between CYP2C19 gene polymorphisms and hypertension in the Hakka population. Loss-of-function genotypes of CYP2C19 increase the risk of hypertension.
BMC CARDIOVASCULAR DISORDERS
(2023)
Article
Medicine, Research & Experimental
Lisa Schweizer, Tina Schaller, Maximilian Zwiebel, Oezge Karayel, Johannes Bruno Mueller-Reif, Wen-Feng Zeng, Sebastian Dintner, Thierry M. Nordmann, Klaus Hirschbuehl, Bruno Maerkl, Rainer Claus, Matthias Mann
Summary: SARS-CoV-2 can cause damage to lung tissue and other organs in the human body, and this study aimed to analyze these effects comprehensively. Using a mass spectrometry proteomics workflow, the researchers identified inflammatory responses as the initial reaction in all tissues. They also found specific patterns of damage in different organs, such as diffuse alveolar damage in the lungs and organ-specific changes in the kidneys, liver, and lymphatic and vascular systems. In the brain, secondary inflammatory effects were linked to neurotransmitter receptors and myelin degradation. These findings contribute to our understanding of the mechanisms of COVID-19 and provide insights for organ-specific therapeutic interventions.
EMBO MOLECULAR MEDICINE
(2023)
Article
Biochemistry & Molecular Biology
Marvin Thielert, Ericka C. M. Itang, Constantin Ammar, Florian A. Rosenberger, Isabell Bludau, Lisa Schweizer, Thierry M. Nordmann, Patricia Skowronek, Maria Wahle, Wen-Feng Zeng, Xie-Xuan Zhou, Andreas-David Brunner, Sabrina Richter, Mitchell P. Levesque, Fabian J. Theis, Martin Steger, Matthias Mann
Summary: Single-cell proteomics allows unbiased characterization of biological function and heterogeneity at the protein level. However, current limitations include proteomic depth, throughput, and robustness. In this study, we introduce a streamlined multiplexed workflow using mDIA to address these limitations. Our approach enables automated and complete dimethyl labeling of bulk or single-cell samples, without compromising proteomic depth. We also demonstrate the ability to quantify twice as many proteins per single cell compared to previous methods, and our workflow allows routine analysis of 80 single cells per day. Additionally, we combine mDIA with spatial proteomics to increase the throughput for microdissection and MS analysis, and successfully identify proteomic signatures of cells within distinct tumor microenvironments in primary cutaneous melanoma.
MOLECULAR SYSTEMS BIOLOGY
(2023)
Article
Engineering, Electrical & Electronic
Qingxin Chu, Wenfeng Zeng
Summary: This paper focuses on the antenna coupling within the MIMO system in 5G and summarizes decoupling techniques. It also elaborates on the new decoupling research developments based on the theory of characteristic mode and provides design examples to validate the proposed decoupling method.
CHINESE JOURNAL OF ELECTRONICS
(2022)