Article
Immunology
Ryan J. Farr, Nathan Godde, Christopher Cowled, Vinod Sundaramoorthy, Diane Green, Cameron Stewart, John Bingham, Carmel M. O'Brien, Megan Dearnley
Summary: Despite being vaccine preventable, rabies still has a significant impact on global mortality, disproportionately affecting children under 15 years of age. Studies have identified 6 cellular miRNAs and 4 exosomal miRNAs that can serve as indicators of lyssavirus infection, providing a foundation for the development of next-generation molecular diagnostics.
FRONTIERS IN CELLULAR AND INFECTION MICROBIOLOGY
(2021)
Article
Engineering, Electrical & Electronic
Nicholas Theobald, Donald Ledvina, Kaitlyn Kukula, Sofia Maines, Kathryn Hasz, Markus Raschke, Jerry Crawford, Jeff Jessing, Yiyan Li
Summary: This study investigates the use of Raman spectroscopy and deep learning techniques for identification of nanofabrication chemicals, demonstrating accurate identification using a neural network without preprocessing. This research has significance for challenging identification applications.
IEEE SENSORS JOURNAL
(2023)
Article
Spectroscopy
Yiming Liu, Ziqi Wang, Zhehai Zhou, Tao Xiong
Summary: This study used single-cell Raman spectroscopy for blood classification and compared different machine learning algorithms. The results showed that support vector machines and artificial neural networks were the most suitable algorithms for single-cell Raman spectrum blood classification, providing essential guidance for future research.
SPECTROCHIMICA ACTA PART A-MOLECULAR AND BIOMOLECULAR SPECTROSCOPY
(2022)
Article
Chemistry, Analytical
Jiazheng Sun, Xuefang Xu, Songsong Feng, Hanyu Zhang, Lingfeng Xu, Hong Jiang, Baibing Sun, Yuyan Meng, Weizhou Chen
Summary: Foodborne illness is a significant global public health problem, and Salmonella infection is a common cause. This study used Raman spectroscopy and convolutional neural networks to rapidly and accurately identify Salmonella serotypes. The results showed that using Savitzky-Golay smoothing combined with Standard Normal Variate preprocessing method achieved the highest accuracy when processing Raman spectral data.
Article
Cell Biology
Concetta Esposito, Mohammed Janneh, Sara Spaziani, Vincenzo Calcagno, Mario Luca Bernardi, Martina Iammarino, Chiara Verdone, Maria Tagliamonte, Luigi Buonaguro, Marco Pisco, Lerina Aversano, Andrea Cusano
Summary: In this study, we investigated the possibility of using Raman spectroscopy assisted by artificial intelligence methods to identify and distinguish liver cancer cells from non-tumor cells. By analyzing the morphological and spectral data of liver cell samples and utilizing machine learning approaches, the study demonstrates the effectiveness of artificial intelligence-assisted Raman spectroscopy in tumor cell classification and prediction, achieving nearly 90% accuracy.
Review
Pharmacology & Pharmacy
Ming Gao, Sibo Liu, Jianan Chen, Keith C. Gordon, Fang Tian, Cushla M. McGoverin
Summary: Drug development is a time-consuming process with high failure rates, where pharmaceutical formulation development plays a crucial role in linking new chemical entities to clinical trials. Artificial intelligence and Raman spectroscopy have the potential to accelerate formulation development and provide new pathways for high-quality data gathering.
INTERNATIONAL JOURNAL OF PHARMACEUTICS
(2021)
Article
Spectroscopy
Maria Gabriela Fernandez-Manteca, Alain A. Ocampo-Sosa, Carlos Ruiz de Alegria-Puig, Maria Pia Roiz, Jorge Rodriguez-Grande, Fidel Madrazo, Jorge Calvo, Luis Rodriguez-Cobo, Jose Miguel Lopez-Higuera, Maria Carmen Farinas, Adolfo Cobo
Summary: In this study, Raman spectroscopy combined with machine learning algorithms was used to explore the automatic identification of eleven species of Candida. The findings showed that a one-dimensional convolutional neural network achieved accurate identification of these pathogenic yeasts.
SPECTROCHIMICA ACTA PART A-MOLECULAR AND BIOMOLECULAR SPECTROSCOPY
(2023)
Article
Chemistry, Analytical
Shuaishuai Yan, Shuying Wang, Jingxuan Qiu, Menghua Li, Dezhi Li, Dongpo Xu, Daixi Li, Qing Liu
Summary: Raman spectroscopy combined with machine learning was used to analyze single-cell spectra of common bacterial strains, achieving efficient prediction of food-borne pathogens at the serotype level.
Article
Biochemical Research Methods
Jessica Zahn, Arno Germond, Alice Y. Lundgren, Marcus T. Cicerone
Summary: This study compared the performance of two machine learning models in discriminating between Raman spectra of different bacterial strains of Escherichia coli. Both models were found to accurately distinguish between the strains, but with different classification boundaries. By analyzing the strain-specific spectral features used by the models, it was observed that they mainly utilized a small subset of high intensity peaks, with only one model utilizing lower intensity peaks.
JOURNAL OF BIOPHOTONICS
(2022)
Article
Microbiology
Jia-Wei Tang, Qing-Hua Liu, Xiao-Cong Yin, Ya-Cheng Pan, Peng-Bo Wen, Xin Liu, Xing-Xing Kang, Bing Gu, Zuo-Bin Zhu, Liang Wang
Summary: Raman spectroscopy (RS) and surface enhanced Raman spectroscopy (SERS) techniques provide potential for fast, sensitive, label-free microbial detection and identification. Different machine learning methods show varying capacities in rapid bacteria differentiation and accurate prediction, with DBSCAN performing best in clustering and CNN as the top model for predicting bacterial species via SERS spectra.
FRONTIERS IN MICROBIOLOGY
(2021)
Article
Multidisciplinary Sciences
Jiarong Ye, Yin-Ting Yeh, Yuan Xue, Ziyang Wang, Na Zhang, He Liu, Kunyan Zhang, RyeAnne Ricker, Zhuohang Yu, Allison Roder, Nestor Perea Lopez, Lindsey Organtini, Wallace Greene, Susan Hafenstein, Huaguang Lu, Elodie Ghedin, Mauricio Terrones, Shengxi Huang, Sharon Xiaolei Huang
Summary: Rapid identification of newly emerging or circulating viruses is crucial for managing public health responses. A portable virus capture device paired with label-free Raman spectroscopy enables fast detection and recognition of viruses using a machine learning approach. A convolutional neural network classifier achieves high accuracy in analyzing Raman spectra of human and avian viruses.
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
(2022)
Article
Multidisciplinary Sciences
Jiarong Ye, Yin-Ting Yeh, Yuan Xue, Ziyang Wang, Na Zhang, He Liu, Kunyan Zhang, RyeAnne Ricker, Zhuohang Yu, Allison Roder, Nestor Perea Lopez, Lindsey Organtini, Wallace Greene, Susan Hafenstein, Huaguang Lu, Elodie Ghedin, Mauricio Terrones, Shengxi Huang, Sharon Xiaolei Huang
Summary: Rapid identification of newly emerging or circulating viruses is crucial for public health response. A study presents a machine learning approach coupled with label-free Raman spectroscopy for fast virus detection and identification. The method achieves high accuracy for various virus type or subtype identification tasks and highlights important Raman spectral ranges for virus identification. The study verifies that the machine learning model effectively recognizes the Raman signatures of proteins, lipids, and other functional groups in different viruses.
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
(2022)
Article
Cell Biology
Xin Huang, Ling Hong, Yuanyuan Wu, Miaoxin Chen, Pengcheng Kong, Jingling Ruan, Xiaoming Teng, Zhiyun Wei
Summary: This study evaluated the potential of follicular fluid (FF) Raman spectra in predicting embryo development and pregnancy outcome for PCOS patients. Specific Raman bands in PCOS FF were identified to have biomarker potential for predicting oocyte developmental potential and clinical pregnancy. Machine-learning algorithms achieved high accuracies in correctly assigning oocyte developmental potential and clinical pregnancy based on Raman spectra of PCOS FF.
FRONTIERS IN CELL AND DEVELOPMENTAL BIOLOGY
(2021)
Article
Spectroscopy
Celestine P. Lawrence
Summary: Recently, a deep convolutional neural network was used to detect liver cancer using Ramanomics. The results showed that an accuracy of about 90% could be achieved after approximately one hour of training on a modern desktop. However, based on my experience with another Ramanomics dataset, simple methods such as logistic regression and random decision forest outperformed deep learning, achieving accuracies of around 90.4% in under a minute and 92.6% in under 10 seconds, respectively. Therefore, while deep learning holds promise, it has yet to provide a significant leap in performance for Ramanomics, highlighting the need for a biophysics aware machine learning approach.
JOURNAL OF RAMAN SPECTROSCOPY
(2023)
Article
Spectroscopy
Ya-Juan Liu, Michelle Kyne, Shuang Wang, Sheng Wang, Xi-Yong Yu, Cheng Wang
Summary: The optimization of Raman instruments has expanded our understanding of single-cell Raman spectroscopy. By improving the speed and sensitivity, as well as implementing advanced data mining methods, complex chemical and biological information within Raman spectral data can be revealed. This paper introduces a new Matlab Graphical User-Friendly Interface (GUI) called CELL IMAGE, which enables the analysis of cellular Raman spectroscopy data. The GUI includes three main steps: spectral processing, pattern recognition, and model validation, and offers various well-known methods for each step. A new subsampling optimization method is integrated into the GUI to estimate the minimum number of spectral collection points. The incorporation of the signal-to-noise ratio (SNR) in the binomial statistical model makes the new subsampling model more sophisticated and suitable for complicated Raman cell data. These embedded methods allow CELL IMAGE to convert spectral information into biological information, including single-cell visualization, cell classification, and biomolecular/drug quantification.
SPECTROCHIMICA ACTA PART A-MOLECULAR AND BIOMOLECULAR SPECTROSCOPY
(2022)
Article
Microbiology
Weiming Tu, Wei E. Huang
ENVIRONMENTAL MICROBIOLOGY
(2023)
Article
Chemistry, Analytical
Jingkai Wang, Siyu Meng, Kaicheng Lin, Xiaofei Yi, Yixiang Sun, Xiaogang Xu, Na He, Zhiqiang Zhang, Huijie Hu, Xingwang Qie, Dayi Zhang, Yuguo Tang, Wei E. Huang, Jian He, Yizhi Song
Summary: A novel culture-independent phenotyping method based on single-cell Raman spectroscopy was proposed for rapid discrimination between fungal and bacterial infections. Three Raman biomarkers were identified for distinguishing yeast and bacterial pathogens. A two-step protocol combining these biomarkers achieved an overall accuracy of 94.9% in differentiating fungal infections from bacterial infections. The method was able to identify fungi in urinary tract infection samples within half an hour and has the potential to be adopted in routine clinical practice.
ANALYTICA CHIMICA ACTA
(2023)
Article
Spectroscopy
Xia Huang, Chengjian Li, Lei Li, Zhiliang Lu, Lihao Zhang, Wei E. Huang, Liang Zhao
Summary: The Raman Deuterium Isotope Probing (Raman-DIP) method was used to study the effects of the GSK2334470 drug on human breast adenocarcinoma cells (MCF-7 cells) in vitro. The cytotoxicity of the drug was evaluated using the Raman-DIP method and CCK-8 assay, and both methods determined that a concentration of 10 μM was cytotoxic and inhibited cell regeneration. The Raman-DIP method accurately monitored the inhibition of cell proliferation by GSK2334470, and combining the results with fingerprint range information improved the evaluation of the drug's sensitivity.
JOURNAL OF RAMAN SPECTROSCOPY
(2023)
Article
Cell Biology
Jie Yang, Chia-Chen Hsu, Ting-Ting Cao, Hua Ye, Jing Chen, Yun-Qing Li
Summary: A hyaluronic acid granular hydrogel was found to promote neuronal and astrocyte colony formation and axonal extension in vitro, mimicking an extracellular matrix structure for neural regeneration. Transplantation of the hydrogel nerve guidance conduit successfully repaired a 10-mm long sciatic nerve gap, showing similar regeneration of axons and myelin sheath, as well as recovery of electrophysiological and motor functions. The conduit outperformed bulk hydrogel or silicone tube transplant. These findings suggest that tissue-engineered nerve conduits with hyaluronic acid granular hydrogels effectively promote the morphological and functional recovery of injured peripheral nerves.
NEURAL REGENERATION RESEARCH
(2023)
Article
Oncology
Tong Yu, Archana Chandrabhan Jadhav, Jiabao B. Xu, Adrian L. L. Harris, Venugopal Nair, Wei E. E. Huang
Summary: Understanding the development of cancer resistance to therapies is crucial. Newcastle disease virus (NDV) is a promising oncolytic agent, but some colon cancer cells show resistance to NDV reinfection. By using Raman spectroscopic and stable isotopic techniques, we found that the resistant cells slow down their replication and divert energy to protein and lipid synthesis. Understanding metabolic reprogramming could aid in developing precision cancer treatments that target resistant cells at the single-cell level.
Article
Biotechnology & Applied Microbiology
Timothee Baudequin, Hazel Wee, Zhanfeng Cui, Hua Ye
Summary: Micro-carriers, such as Ca-alginate and Fe-alginate, have been studied as substrates for mesenchymal stem cell (MSCs) proliferation. While Ca-alginate requires functionalization, Fe-alginate beads showed great potential as ready-to-use carriers for MSCs without the need for additional modifications.
BIOENGINEERING-BASEL
(2023)
Review
Biotechnology & Applied Microbiology
Jochem R. Nielsen, Ruud A. Weusthuis, Wei E. Huang
Summary: Enzymes in commercial bioproduction require high efficiency, robustness, and specificity. Enzyme engineering techniques like random mutagenesis and directed evolution are often used to achieve these properties. Growth-coupling selection strategies can be used to select enzyme variants based on improved cofactor oxidation or reduction rates. This review summarizes the metabolic engineering involved in creating strains auxotrophic for the oxidized or reduced state of redox cofactors and highlights the successful applications of this technique in enzyme engineering.
BIOTECHNOLOGY ADVANCES
(2023)
Article
Endocrinology & Metabolism
David Habart, Adam Koza, Ivan Leontovyc, Lucie Kosinova, Zuzana Berkova, Jan Kriz, Klara Zacharovova, Bas Brinkhof, Dirk-Jan Cornelissen, Nicholas Magrane, Katerina Bittenglova, Martin Capek, Jan Valecka, Alena Habartova, Frantisek Saudek
Summary: We developed IsletSwipe, a platform consisting of a web interface and a mobile application, for experts to exchange graphical opinions and facilitate consensus formation in deep learning-based islet counting. The platform was tested in a pilot study with nine experts from three centers. The study demonstrated the functionalities and use case scenarios, and showed that IsletSwipe is a suitable tool for consensus finding.
Article
Polymer Science
Guoying Zhou, Jiayan Zhu, Catriona Inverarity, Yifeng Fang, Zhao Zhang, Hua Ye, Zhanfeng Cui, Linh Nguyen, Haitong Wan, Julian F. Dye
Summary: A fibrin/PVA scaffold with interconnected porous structures and preserved fibrous architecture was fabricated using an emulsion templating method. The scaffold showed excellent biocompatibility and efficacy for dermal reconstruction, promoting wound healing and tissue regeneration.
Article
Microbiology
Jiabao Xu, Yanjun Luo, Jingkai Wang, Weiming Tu, Xiaofei Yi, Xiaogang Xu, Yizhi Song, Yuguo Tang, Xiaoting Hua, Yunsong Yu, Huabing Yin, Qiwen Yang, Wei E. Huang
Summary: Integrating AI and new diagnostic platforms in clinical microbiology laboratories can improve efficiency by reducing turnaround time and cost. Using single-cell Raman spectroscopy and AI, rapid identification of infectious fungi with high accuracy was achieved.
FRONTIERS IN MICROBIOLOGY
(2023)
Article
Environmental Sciences
Tingting Teng, Wei E. Huang, Guanghe Li, Xinzi Wang, Yizhi Song, Xiaoyi Tang, Dunzhu Dawa, Bo Jiang, Dayi Zhang
Summary: In this study, an assembly technology for manufacturing Acinetobacter-based biosensor arrays using magnetic nanoparticle functionalization was developed to solve the low throughput and complicated operation problems. The biosensor array showed high viability, sensitivity, and specificity in sensing multiple contaminants in a high-throughput manner, and remained acceptable for at least 20 days. Positive correlations were found between the biosensor estimation and chemical analysis, indicating the feasibility of this magnetic nanoparticle-functionalized biosensor array for online environmental monitoring at contaminated sites.
SCIENCE OF THE TOTAL ENVIRONMENT
(2023)
Article
Engineering, Environmental
Menghan Wu, Xinning Liu, Weiming Tu, Juntao Xia, Yina Zou, Xiaoqiang Gong, Peng Yu, Wei E. Huang, Hui Wang
Summary: In this study, a new perspective on the rationality of oriented propionate production using mixed microbial cultures (MMCs) with lactate as feedstock was proposed. The feasibility of the food waste-lactate-propionate route to reverse the original butyrate-type fermentation situation and mechanisms for maintaining stability were verified.
Article
Chemistry, Multidisciplinary
Jiabao Xu, Tiffany Lodge, Caroline Kingdon, James W. L. Strong, John Maclennan, Eliana Lacerda, Slawomir Kujawski, Pawel Zalewski, Wei E. Huang, Karl J. Morten
Summary: This study demonstrates the potential of using a single-cell Raman platform and artificial intelligence to diagnose and manage ME/CFS. The Raman profiles of blood cells can accurately distinguish between healthy individuals, disease controls, and ME/CFS patients, as well as differentiate between different severity levels of ME/CFS. The identification of specific Raman peaks also provides insights into biological changes and potential therapeutic targets.
Article
Physics, Multidisciplinary
Casey Adam, Celine Kayal, Ari Ercole, Sonia Contera, Hua Ye, Antoine Jerusalem
Summary: This study demonstrates that general anaesthetics affect the viscoelasticity and functional activity of cells simultaneously, and that the alterations in firing and viscoelasticity are correlated.
COMMUNICATIONS PHYSICS
(2023)
Article
Biotechnology & Applied Microbiology
Carla V. Fuenteslopez, Mark S. Thompson, Hua Ye
Summary: Traumatic injuries are a significant global health issue, but there is limited research on microvascular traumatic injuries. This study aims to develop an in vitro model for studying traumatic injuries at the microvascular level. Different hydrogels were created using various polymers, solvents, and concentrations, and evaluated for their effects on cell behavior and hydrogel properties. Fibrin hydrogels at 3% and 5% w/v in serum-free media were found to be the most suitable for further experimentation, enabling the formation of interconnected capillary-like networks.
BIOENGINEERING-BASEL
(2023)