Automating FACS data analysis for HIV studies

描述（由申请人提供）：流式细胞术（FACS）技术的最新进展为理解、监测和对抗HIV疾病提供了新的和令人兴奋的方法。多参数FACS（4色，2散射）仪器测量4至6种荧光颜色，已被从事艾滋病毒疫苗开发和关键艾滋病毒研究的前沿实验室常规使用。现在进入这一领域的高维FACS仪器通过在单个细胞上或细胞内提供多达十几种或更多荧光颜色的高质量测量，扩展了这种能力。此外，结合细胞内细胞因子、转录因子、磷酸化蛋白和病毒衍生蛋白检测记忆和其他T细胞表面标记物的染色方法，现在可以测量更精细划分的T细胞亚群的功能反应；而且，新的数据分析方法现在可以更好地可视化高分辨率FACS数据并从数据中提取结果。这种技术的大部分已经在开发新的检测方法以监测艾滋病毒疫苗有效性的实验室中得到应用。然而，新的（和旧的）FACS技术的广泛使用受到学习如何有效使用该技术所需的时间和精力的严重限制，重要的是，如何将该技术常规应用于分析在测试候选艾滋病毒疫苗和其他类型的艾滋病毒临床研究中遇到的大量样本的问题尚未解决。这里提出的研究旨在解决这些问题，这些问题对研制艾滋病毒疫苗取得进展尤为重要。我们已经开发了基于知识的协议设计工具，可以加快FACS染色协议的开发和实施，促进FACS数据收集，并永久存储FACS数据分析所需的协议信息。此外，我们已经设计了无需干预即可操作的方法/软件，以标准化Hi-D FACS数据并纠正（补偿）荧光重叠。最后，我们已经实现了新的聚类算法的原型，这些算法在统计上适用于分析FACS数据，即，与微阵列和其他聚类算法不同，我们的算法适用于包含相对较少的测量（荧光颜色等）的数据集，这些数据集用于非常大量的项目（细胞或其他颗粒）。在这里，我们建议完成并链接这些软件组件，以创建一个工作原型，该原型将通过开发软件来支持HIV疫苗的开发，该软件将1)自动准备用于分析的FACS数据（数据标准化和荧光补偿）；2)促进FACS数据分析（“聚类”以提供子集识别和门控）；3)临床疫苗试验中测试样本和参考样本数据的统计比较自动化；从而促进和提高在研究和临床研究中评价艾滋病毒疫苗反应的能力。

DESCRIPTION (provided by applicant): Recent advances in flow cytometry (FACS) technology offer new and exciting approaches to understanding, monitoring and combating HIV disease. Multiparameter FACS (4 color, 2 scatter) instruments that measure four to six fluorescence colors are already routinely used by forefront laboratories doing HIV vaccine development and key HIV research. High-dimensional FACS instruments now coming into this arena extend this capability by providing high-quality measurements for up to a dozen or more fluorescence colors on or in individual cells. In addition, staining methods that detect memory and other T cell surface markers in combination with intracellular cytokines, transcription factors, phosphoproteins and virally-derived proteins now enable measurement of functional responses in ever more finely divided T cell subsets; and, new data analysis methods now enable better visualization of Hi-D FACS data and extraction of findings from that data. Much of this technology is already being applied in laboratories developing new assays to monitor HIV vaccine effectiveness. However, broad utilization of the new (and the older) FACS technology is sharply limited both by the time and effort required to learn how to effectively use the technology and, importantly, by the as yet unsolved problem of how to routinely apply the technology to assaying the large numbers of samples encountered in testing candidate HIV vaccines and other kinds of HIV clinical research. Studies proposed here address these issues, which are particularly crucial to progress in the development of HIV vaccines. We have already developed knowledge-based protocol design tools that speed up and the development and implementation of FACS staining protocols, facilitate FACS data collection, and permanently store protocol information necessary for FACS data analysis. In addition, we have already designed methods/software that can operate without intervention to standardize Hi-D FACS data and correct (compensate) for fluorescence overlaps. Finally, we have already implemented prototypes of new clustering algorithms that are statistically appropriate for analyzing FACS data, i.e., unlike microarray and other clustering algorithms, our algorithms work for data sets that contain relatively few measurements (fluorescence colors, etc.) taken for a very large number of items (cells or other particles). Here, we propose to complete and link these software components to create a working prototype that will support HIV vaccine development by developing software that will 1) automate preparation of FACS data for analysis (data standardization and fluorescence compensation); 2) facilitate FACS data analysis ("clustering" to provide subset identification and gating); 3) automate of statistical comparisons of data from test and reference samples in clinical vaccine trials; and thereby, 4) facilitate and improve the ability to evaluate HIV vaccine responses in research and clinical studies.