arXiv - QuanBio - Quantitative Methods最新文献_第4页

Focusing Viral Risk Ranking Tool on Prediction 将病毒风险排名工具的重点放在预测上

arXiv - QuanBio - Quantitative Methods Pub Date : 2024-09-07 DOI: arxiv-2409.04932

Katherine Budeski, Marc Lipsitch

{"title":"Focusing Viral Risk Ranking Tool on Prediction","authors":"Katherine Budeski, Marc Lipsitch","doi":"arxiv-2409.04932","DOIUrl":"https://doi.org/arxiv-2409.04932","url":null,"abstract":"Preparing to rapidly respond to emerging infectious diseases is becoming ever\u0000more critical. \"SpillOver: Viral Risk Ranking\" is an open-source tool developed\u0000to evaluate novel wildlife-origin viruses for their risk of spillover from\u0000animals to humans and their risk of spreading in human populations. However,\u0000several of the factors used in the risk assessment are dependent on evidence of\u0000previous zoonotic spillover and/or sustained transmission in humans. Therefore,\u0000we performed a reanalysis of the \"Ranking Comparison\" after removing eight\u0000factors that require post-spillover knowledge and compared the adjusted risk\u0000rankings to the originals. The top 10 viruses as ranked by their adjusted\u0000scores also had very high original scores. However, the predictive power of the\u0000tool for whether a virus was a human virus or not as classified in the\u0000Spillover database deteriorated when these eight factors were removed. The area\u0000under the receiver operating characteristic curves (AUROC) for the original\u0000score, 0.94, decreased to 0.73 for the adjusted scores. Furthermore, we\u0000compared the mean and standard deviation of the human and non-human viruses at\u0000the factor level. Most of the excluded spillover-dependent factors had\u0000dissimilar means between the human and non-human virus groups compared to the\u0000non-spillover dependent factors, which frequently demonstrated similar means\u0000between the two groups with some exceptions. We concluded that the original\u0000formulation of the tool depended heavily on spillover-dependent factors to\u0000\"predict\" the risk of zoonotic spillover for a novel virus. Future iterations\u0000of the tool should take into consideration other non-spillover dependent\u0000factors and omit those that are spillover-dependent to ensure the tool is fit\u0000for purpose.","PeriodicalId":501266,"journal":{"name":"arXiv - QuanBio - Quantitative Methods","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142213280","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Unlocking Potential Binders: Multimodal Pretraining DEL-Fusion for Denoising DNA-Encoded Libraries 释放潜在粘合剂：用于去噪 DNA 编码文库的多模式预训练 DEL-Fusion

arXiv - QuanBio - Quantitative Methods Pub Date : 2024-09-07 DOI: arxiv-2409.05916

Chunbin Gu, Mutian He, Hanqun Cao, Guangyong Chen, Chang-yu Hsieh, Pheng Ann Heng

{"title":"Unlocking Potential Binders: Multimodal Pretraining DEL-Fusion for Denoising DNA-Encoded Libraries","authors":"Chunbin Gu, Mutian He, Hanqun Cao, Guangyong Chen, Chang-yu Hsieh, Pheng Ann Heng","doi":"arxiv-2409.05916","DOIUrl":"https://doi.org/arxiv-2409.05916","url":null,"abstract":"In the realm of drug discovery, DNA-encoded library (DEL) screening\u0000technology has emerged as an efficient method for identifying high-affinity\u0000compounds. However, DEL screening faces a significant challenge: noise arising\u0000from nonspecific interactions within complex biological systems. Neural\u0000networks trained on DEL libraries have been employed to extract compound\u0000features, aiming to denoise the data and uncover potential binders to the\u0000desired therapeutic target. Nevertheless, the inherent structure of DEL,\u0000constrained by the limited diversity of building blocks, impacts the\u0000performance of compound encoders. Moreover, existing methods only capture\u0000compound features at a single level, further limiting the effectiveness of the\u0000denoising strategy. To mitigate these issues, we propose a Multimodal\u0000Pretraining DEL-Fusion model (MPDF) that enhances encoder capabilities through\u0000pretraining and integrates compound features across various scales. We develop\u0000pretraining tasks applying contrastive objectives between different compound\u0000representations and their text descriptions, enhancing the compound encoders'\u0000ability to acquire generic features. Furthermore, we propose a novel DEL-fusion\u0000framework that amalgamates compound information at the atomic, submolecular,\u0000and molecular levels, as captured by various compound encoders. The synergy of\u0000these innovations equips MPDF with enriched, multi-scale features, enabling\u0000comprehensive downstream denoising. Evaluated on three DEL datasets, MPDF\u0000demonstrates superior performance in data processing and analysis for\u0000validation tasks. Notably, MPDF offers novel insights into identifying\u0000high-affinity molecules, paving the way for improved DEL utility in drug\u0000discovery.","PeriodicalId":501266,"journal":{"name":"arXiv - QuanBio - Quantitative Methods","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142213268","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Large Language Models in Drug Discovery and Development: From Disease Mechanisms to Clinical Trials 药物发现和开发中的大型语言模型：从疾病机理到临床试验

arXiv - QuanBio - Quantitative Methods Pub Date : 2024-09-06 DOI: arxiv-2409.04481

Yizhen Zheng, Huan Yee Koh, Maddie Yang, Li Li, Lauren T. May, Geoffrey I. Webb, Shirui Pan, George Church

引用次数: 0

The Effects of Unilateral Slope Loading on Lower Limb Plantar Flexor Muscle EMG Signals in Young Healthy Males 单侧斜坡负荷对健康男性下肢跖屈肌肌电信号的影响

arXiv - QuanBio - Quantitative Methods Pub Date : 2024-09-06 DOI: arxiv-2409.04321

Xinyu Zhou, Gengshang Dong, Pengxuan Zhang

{"title":"The Effects of Unilateral Slope Loading on Lower Limb Plantar Flexor Muscle EMG Signals in Young Healthy Males","authors":"Xinyu Zhou, Gengshang Dong, Pengxuan Zhang","doi":"arxiv-2409.04321","DOIUrl":"https://doi.org/arxiv-2409.04321","url":null,"abstract":"Different loading modes can significantly affect human gait, posture, and\u0000lower limb biomechanics. This study investigated the muscle activity intensity\u0000of the lower limb soleus muscle in the slope environment of young healthy adult\u0000male subjects under unilateral loading environment. Ten subjects held dumbbells\u0000equal to 5% and 10% of their body weight (BW) and walked at a fixed speed on a\u0000slope of 5 degree and 10 degree, respectively. The changes of electromyography\u0000(EMG) of bilateral soleus muscles of the lower limbs were recorded. Experiments\u0000were performed using one-way analysis of variance (ANOVA) and multivariate\u0000analysis of variance (MANOVA) to examine the relationship between load weight,\u0000slope angle, and muscle activity intensity. The data provided by this research\u0000can help to promote the development of the field of lower limb assist\u0000exoskeleton. The research results fill the missing data when loading on the\u0000slope side, provide data support for future assistance systems, and promote the\u0000formation of relevant data sets, so as to improve the terrain recognition\u0000ability and the movement ability of the device wearer.","PeriodicalId":501266,"journal":{"name":"arXiv - QuanBio - Quantitative Methods","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142213290","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Protein sequence classification using natural language processing techniques 利用自然语言处理技术进行蛋白质序列分类

arXiv - QuanBio - Quantitative Methods Pub Date : 2024-09-06 DOI: arxiv-2409.04491

Huma PerveenSchool of Mathematical and Physical Sciences, University of Sussex, Brighton, UK, Julie WeedsSchool of Engineering and Informatics, University of Sussex, Brighton, UK

{"title":"Protein sequence classification using natural language processing techniques","authors":"Huma PerveenSchool of Mathematical and Physical Sciences, University of Sussex, Brighton, UK, Julie WeedsSchool of Engineering and Informatics, University of Sussex, Brighton, UK","doi":"arxiv-2409.04491","DOIUrl":"https://doi.org/arxiv-2409.04491","url":null,"abstract":"Proteins are essential to numerous biological functions, with their sequences\u0000determining their roles within organisms. Traditional methods for determining\u0000protein function are time-consuming and labor-intensive. This study addresses\u0000the increasing demand for precise, effective, and automated protein sequence\u0000classification methods by employing natural language processing (NLP)\u0000techniques on a dataset comprising 75 target protein classes. We explored\u0000various machine learning and deep learning models, including K-Nearest\u0000Neighbors (KNN), Multinomial Na\"ive Bayes, Logistic Regression, Multi-Layer\u0000Perceptron (MLP), Decision Tree, Random Forest, XGBoost, Voting and Stacking\u0000classifiers, Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM),\u0000and transformer models (BertForSequenceClassification, DistilBERT, and\u0000ProtBert). Experiments were conducted using amino acid ranges of 1-4 grams for\u0000machine learning models and different sequence lengths for CNN and LSTM models.\u0000The KNN algorithm performed best on tri-gram data with 70.0% accuracy and a\u0000macro F1 score of 63.0%. The Voting classifier achieved best performance with\u000074.0% accuracy and an F1 score of 65.0%, while the Stacking classifier reached\u000075.0% accuracy and an F1 score of 64.0%. ProtBert demonstrated the highest\u0000performance among transformer models, with a accuracy 76.0% and F1 score 61.0%\u0000which is same for all three transformer models. Advanced NLP techniques,\u0000particularly ensemble methods and transformer models, show great potential in\u0000protein classification. Our results demonstrate that ensemble methods,\u0000particularly Voting Soft classifiers, achieved superior results, highlighting\u0000the importance of sufficient training data and addressing sequence similarity\u0000across different classes.","PeriodicalId":501266,"journal":{"name":"arXiv - QuanBio - Quantitative Methods","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142213284","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

The Role of Graph Topology in the Performance of Biomedical Knowledge Graph Completion Models 图拓扑在生物医学知识图完成模型性能中的作用

arXiv - QuanBio - Quantitative Methods Pub Date : 2024-09-06 DOI: arxiv-2409.04103

Alberto Cattaneo, Stephen Bonner, Thomas Martynec, Carlo Luschi, Ian P Barrett, Daniel Justus

引用次数: 0

How hosts and pathogens choose the strengths of defense and counter-defense. A game-theoretical view 宿主和病原体如何选择防御和反防御的强度？博弈论观点

arXiv - QuanBio - Quantitative Methods Pub Date : 2024-09-06 DOI: arxiv-2409.04497

Shalu Dwivedi, Ravindra Garde, Stefan Schuster

{"title":"How hosts and pathogens choose the strengths of defense and counter-defense. A game-theoretical view","authors":"Shalu Dwivedi, Ravindra Garde, Stefan Schuster","doi":"arxiv-2409.04497","DOIUrl":"https://doi.org/arxiv-2409.04497","url":null,"abstract":"Host-pathogen interactions consist of an attack by the pathogen, frequently a\u0000defense by the host and possibly a counter-defense by the pathogen. Here, we\u0000present a game-theoretical approach to describing such interactions. We\u0000consider a game where the host and pathogen are players and they can choose\u0000between the strategies of defense (or counter-defense) and no response.\u0000Specifically, they may or may not produce a toxin and an enzyme degrading the\u0000toxin, respectively. We consider that the host and pathogen must also incur a\u0000cost for toxin or enzyme production. We highlight both the sequential and\u0000non-sequential versions of the game and determine the Nash equilibria. Further,\u0000we resolve a paradox occurring in that interplay. If the inactivating enzyme is\u0000very efficient, producing the toxin becomes useless, leading to the enzyme\u0000being no longer required. Then, production of the defense becomes useful again.\u0000In game theory, such situations can be described by a generalized matching\u0000pennies game. As a novel result, we find under which conditions the defense\u0000cycle leads to a steady state or to an oscillation. We obtain, for saturating\u0000dose-response kinetics and considering monotonic cost functions, 'partial\u0000(counter-)defense' strategies as pure Nash equilibria. This implies that\u0000producing a moderate amount of toxin and enzyme is the best choice.","PeriodicalId":501266,"journal":{"name":"arXiv - QuanBio - Quantitative Methods","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142213283","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Enhancing Clinical Data Warehouses with Provenance and Large File Management: The gitOmmix Approach for Clinical Omics Data 利用出处和大文件管理增强临床数据仓库：临床奥米克斯数据的 gitOmmix 方法

arXiv - QuanBio - Quantitative Methods Pub Date : 2024-09-05 DOI: arxiv-2409.03288

Maxime WackCRC, HeKA, HEGP, CHNO, Adrien CouletCRC, HeKA, Anita BurgunHEGP, Imagine, Bastien RanceUPCité, HEGP, CRC, HeKA

{"title":"Enhancing Clinical Data Warehouses with Provenance and Large File Management: The gitOmmix Approach for Clinical Omics Data","authors":"Maxime WackCRC, HeKA, HEGP, CHNO, Adrien CouletCRC, HeKA, Anita BurgunHEGP, Imagine, Bastien RanceUPCité, HEGP, CRC, HeKA","doi":"arxiv-2409.03288","DOIUrl":"https://doi.org/arxiv-2409.03288","url":null,"abstract":"Background. Clinical data warehouses (CDWs) are essential in the reuse of\u0000hospital data in observational studies or predictive modeling. However, state\u0000of-the-art CDW systems present two drawbacks. First, they do not support the\u0000management of large data files, what is critical in medical genomics,\u0000radiology, digital pathology, and other domains where such files are generated.\u0000Second, they do not provide provenance management or means to represent\u0000longitudinal relationships between patient events. Indeed, a disease diagnosis\u0000and its follow-up rely on multiple analyses. In these cases no relationship\u0000between the data (e.g., a large file) and its associated analysis and decision\u0000can be documented.Method. We introduce gitOmmix, an approach that overcomes\u0000these limitations, and illustrate its usefulness in the management of medical\u0000omics data. gitOmmix relies on (i) a file versioning system: git, (ii) an\u0000extension that handles large files: git-annex, (iii) a provenance knowledge\u0000graph: PROV-O, and (iv) an alignment between the git versioning information and\u0000the provenance knowledge graph.Results. Capabilities inherited from git and\u0000git-annex enable retracing the history of a clinical interpretation back to the\u0000patient sample, through supporting data and analyses. In addition, the\u0000provenance knowledge graph, aligned with the git versioning information,\u0000enables querying and browsing provenance relationships between these\u0000elements.Conclusion. gitOmmix adds a provenance layer to CDWs, while scaling to\u0000large files and being agnostic of the CDW system. For these reasons, we think\u0000that it is a viable and generalizable solution for omics clinical studies.","PeriodicalId":501266,"journal":{"name":"arXiv - QuanBio - Quantitative Methods","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142213289","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Adaptive cross-country optimisation strategies in thermal soaring birds 热翱翔鸟类的适应性越野优化策略

arXiv - QuanBio - Quantitative Methods Pub Date : 2024-09-05 DOI: arxiv-2409.03849

Göksel Keskin, Olivier Duriez, Pedro Lacerda, Andrea Flack, Máté Nagy

{"title":"Adaptive cross-country optimisation strategies in thermal soaring birds","authors":"Göksel Keskin, Olivier Duriez, Pedro Lacerda, Andrea Flack, Máté Nagy","doi":"arxiv-2409.03849","DOIUrl":"https://doi.org/arxiv-2409.03849","url":null,"abstract":"Thermal soaring enables birds to perform cost-efficient flights during\u0000foraging or migration trips. Yet, although all soaring birds exploit vertical\u0000winds effectively, this group contains species that vary strongly in their\u0000morphologies. Aerodynamic rules dictate the costs and benefits of flight, but,\u0000depending on their ecological needs, species may use different behavioural\u0000strategies. To quantify these morphology-related differences in behavioural\u0000cross-country strategies, we compiled and analysed a large dataset, which\u0000includes data from over a hundred individuals from 12 soaring species recorded\u0000with high frequency tracking devices. We quantified the performance during\u0000thermalling and gliding flights, and the overall cross-country behaviour that\u0000is the combination of both. Our results confirmed aerodynamic theory across the\u000012 species; species with higher wing loading typically flew faster, and\u0000consequently turned on a larger radius, than lighter ones. Furthermore, the\u0000combination of circling radius and minimum sink speed determines the maximum\u0000benefits soaring birds can obtain from thermals. Also, we observed a spectrum\u0000of strategies regarding the adaptivity to thermal strength and uncovered a\u0000universal rule for cross-country strategies for all analysed species. Finally,\u0000our newly described behavioural rules can provide inspirations for technical\u0000applications, like the development of autopilot systems for autonomous robotic\u0000gliders.","PeriodicalId":501266,"journal":{"name":"arXiv - QuanBio - Quantitative Methods","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142213288","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Temporal and Spacial Studies of Infectious Diseases: Mathematical Models and Numerical Solvers 传染病的时空研究：数学模型和数值求解器

arXiv - QuanBio - Quantitative Methods Pub Date : 2024-09-05 DOI: arxiv-2409.10556

Md Abu Talha, Yongjia Xu, Shan Zhao, Weihua Geng

{"title":"Temporal and Spacial Studies of Infectious Diseases: Mathematical Models and Numerical Solvers","authors":"Md Abu Talha, Yongjia Xu, Shan Zhao, Weihua Geng","doi":"arxiv-2409.10556","DOIUrl":"https://doi.org/arxiv-2409.10556","url":null,"abstract":"The SIR model is a classical model characterizing the spreading of infectious\u0000diseases. This model describes the time-dependent quantity changes among\u0000Susceptible, Infectious, and Recovered groups. By introducing space-depend\u0000effects such as diffusion and creation in addition to the SIR model, the\u0000Fisher's model is in fact a more advanced and comprehensive model. However, the\u0000Fisher's model is much less popular than the SIR model in simulating infectious\u0000disease numerically due to the difficulties from the parameter selection, the\u0000involvement of 2-d/3-d spacial effects, the configuration of the boundary\u0000conditions, etc. This paper aim to address these issues by providing numerical algorithms\u0000involving space and time finite difference schemes and iterative methods, and\u0000its open-source Python code for solving the Fisher's model. This 2-D Fisher's\u0000solver is second order in space and up to the second order in time, which is\u0000rigorously verified using test cases with analytical solutions. Numerical\u0000algorithms such as SOR, implicit Euler, Staggered Crank-Nicolson, and ADI are\u0000combined to improve the efficiency and accuracy of the solver. It can handle\u0000various boundary conditions subject to different physical descriptions. In\u0000addition, real-world data of Covid-19 are used by the model to demonstrate its\u0000practical usage in providing prediction and inferences.","PeriodicalId":501266,"journal":{"name":"arXiv - QuanBio - Quantitative Methods","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142264958","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0