CAREER: ABI-Innovation: CiliaWeb: Integrated platform for foundational and reproducible ciliary beat pattern analysis

职业：ABI-创新：CiliaWeb：用于基础和可重复纤毛跳动模式分析的集成平台

• 批准号: 1845915
• 负责人: Shannon Quinn
• 金额: $ 96.41万
• 依托单位: University of Georgia Research Foundation Inc
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-08-01 至 2024-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1845915&HistoricalAwards=false
• 关键词: CAREER; ABI; Innovation; CiliaWeb; Integrated

Cilia are microscopic hairs that protrude from nearly every cell in the human body, including those in the throat, nose, lungs, kidneys, and brain. Cilia move in regular, rhythmic patterns in order to transport materials outside the cells. These movements of the cilia are important at every stage of life, from early embryonic development to reproduction and regular healthy maintenance; in short, healthy cilia are a critical component of overall health of an organism. Consequently, when cilia move abnormally, numerous debilitating conditions can result, encompassing disorders ranging from developmental shortcomings in the embryo to adult pathologies such as infertility and lung scarring. Therefore, the key question this proposal seeks to answer is: by observing the motion of cilia, is it possible to predict whether the morphology or behavior is abnormal, and if so, whether that particular abnormality is associated with a specific condition? The answer to this question has implications not only in human health, but also in building fundamental knowledge of the biology around how cilia are constructed and maintained. To address this question, we are building the CiliaWeb platform, which will include new algorithms for analyzing videos of cilia and shedding light on the statistics of ciliary motion. The algorithms will take advantage of the regular back-and-forth movements of cilia and be built to incorporate crowdsourced feedback from users to improve the models of ciliary motion. As its name implies, CiliaWeb will be internet-accessible for researchers and clinicians to upload datasets, conduct analyses, and visualize results. The ultimate goal of CiliaWeb is to provide a standardized platform for ciliary motion analysis that can be reproduced and validated by others, encouraging collaboration and catalyzing new discoveries in fundamental cilia research as well as connections to health and well-being.Cilia are microscopic hairs that protrude from eukaryotic cells and are critical components in development, reproduction, and homeostasis. Cilia beat in synchronous waves to generate fluid flow and clear particulates; disruptions of these motions result in a spectrum of pathologies. However, no quantitative, reproducible, and validated method exists for assessing ciliary motion from high-speed videomicroscopy. Without such a method, beat pattern phenotype cannot be used to conclusively advance our understanding of the fundamental biology of cilia (construction and maintenance), nor the precise role of beat pattern in organismal health (examples being signaling, development, fertility). This proposal aims to create the CiliaWeb framework to establish an objective measure of motion phenotypes, enable new lines of inquiry into biophysical mechanisms of ciliary motion and discover the biological roles of specific motion phenotypes, as well as make possible cross-institutional collaborations, large-scale genotype-phenotype association studies, and longitudinal biomedical studies. The CiliaWeb platform will incorporate tightly-integrated and openly-available tools for processing ciliary motion, including: a novel unsupervised dynamic texture segmentation algorithm for automatically identifying cilia in high-speed digital videos; a hierarchical encoding scheme for quantifying ciliary motion at multiple spatiotemporal scales; a crowdsourcing module to solicit targeted feedback on the learned motion patterns that are fully interactive and automated. CiliaWeb will contribute methodological advancements in computational bioimaging and establish baseline data through the release of open source code and validated datasets, in addition to hands-on outreach and training for students in the form of workshops and hackathons. Project updates: https://quinngroup.github.ioThis award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

纤毛是一种微小的毛发，几乎从人体的每一个细胞中伸出，包括喉咙，鼻子，肺，肾脏和大脑中的细胞。纤毛以有规律、有节奏的方式移动，以便将物质运输到细胞外。纤毛的这些运动在生命的每个阶段都很重要，从早期胚胎发育到生殖和定期健康维护;简而言之，健康的纤毛是生物体整体健康的关键组成部分。因此，当纤毛运动异常时，可能会导致许多使人衰弱的疾病，包括从胚胎发育缺陷到成年病理学（如不孕症和肺瘢痕）的疾病。因此，这项建议试图回答的关键问题是：通过观察纤毛的运动，是否有可能预测形态或行为是否异常，如果是这样，这种特定的异常是否与特定的条件有关？这个问题的答案不仅对人类健康有影响，而且对建立关于纤毛如何构建和维持的生物学基础知识也有影响。为了解决这个问题，我们正在构建CiliaWeb平台，该平台将包括用于分析纤毛视频和揭示纤毛运动统计数据的新算法。这些算法将利用纤毛的定期来回运动，并将来自用户的众包反馈纳入其中，以改进纤毛运动模型。顾名思义，CiliaWeb将可供研究人员和临床医生通过互联网访问，以上传数据集，进行分析和可视化结果。CiliaWeb的最终目标是为纤毛运动分析提供一个标准化的平台，可以被其他人复制和验证，鼓励合作，促进基础纤毛研究的新发现以及与健康和福祉的联系。纤毛是从真核细胞中伸出的微观毛发，是发育，繁殖和稳态的关键组成部分。纤毛在同步波中跳动以产生流体流动和清除颗粒;这些运动的中断导致一系列病理。然而，没有定量的，可重复的，和验证的方法来评估睫状体运动的高速视频显微镜。如果没有这样的方法，节拍模式表型就不能用于最终推进我们对纤毛基本生物学（构建和维护）的理解，也不能用于节拍模式在生物健康中的确切作用（例如信号传导，发育，生育）。该提案旨在创建CiliaWeb框架，以建立运动表型的客观测量，使新的调查线进入纤毛运动的生物物理机制，并发现特定运动表型的生物学作用，以及使跨机构合作，大规模基因型-表型关联研究和纵向生物医学研究成为可能。CiliaWeb平台将采用紧密集成和开放式的工具来处理纤毛运动，包括：一种新型的无监督动态纹理分割算法，用于自动识别高速数字视频中的纤毛;一种分层编码方案，用于在多个时空尺度上量化纤毛运动;一个众包模块，用于征求关于完全交互和自动化的学习运动模式的有针对性的反馈。CiliaWeb将通过发布开源代码和经验证的数据集，以及以研讨会和黑客马拉松的形式为学生提供实践推广和培训，为计算生物成像的方法学进步做出贡献，并建立基线数据。项目更新：https://quinngroup.github.ioThis奖项反映了NSF的法定使命，并被认为值得通过使用基金会的知识价值和更广泛的影响审查标准进行评估来支持。

项目成果

Variational Autoencoders For Semi-Supervised Deep Metric Learning

• DOI: 10.25080/majora-212e5952-022
• 发表时间: 2022
• 作者: Nathan Safir;Meekail Zain;C. Godwin;Eric L. Miller;Bella Humphrey;Shannon Quinn

Incorporating Task-Agnostic Information in Task-Based Active Learning Using a Variational Autoencoder

• DOI: 10.25080/majora-212e5952-011
• 发表时间: 2022
• 作者: C. Godwin;Meekail Zain;Nathan Safir;Bella Humphrey;Shannon Quinn

Towards an Unsupervised Spatiotemporal Representation of Cilia Video Using A Modular Generative Pipeline

使用模块化生成管道实现纤毛视频的无监督时空表示

• DOI: 10.25080/majora-342d178e-017
• 发表时间: 2020
• 期刊: Proceedings of the Python in Science Conference
• 作者: Zain, Meekail;Rao, Sonia;Safir, Nathan;Wyner, Quinn;Humphrey, Isabella;Eldridge, Alex;Li, Chenxiao;AlAila, BahaaEddin;Quinn, Shannon
• 通讯作者: Quinn, Shannon

Low Level Feature Extraction for Cilia Segmentation

• DOI: 10.25080/majora-212e5952-026
• 发表时间: 2022
• 作者: Meekail Zain;E. Miller;Shannon Quinn;Cecilia W. Lo