Proteomics software package for the detection of cancer

描述（由申请人提供）： 在SBIR项目的第二阶段，我们将开发一个模块化和可扩展的探索性软件工具，用于癌症诊断开发所需的蛋白质组学MS的信号调节和数据挖掘。该软件将允许研究人员构建和验证质谱（MS）数据的迭代分类方法。该方法将被开发用于表面增强激光解吸电离（SELDI）MS，但算法本身和模块化框架将适用于其他类型的MS数据的分析和分类，如基质辅助激光解吸电离（MALDI）和电喷雾电离（ESI）。该分类软件将能够诊断疾病并识别MS记录中的临床显著特征。用户界面将以现有的INCOGEN VIBE软件为基础，开发该软件是为了提供一个方便用户的、可扩展的数据分析框架。拟议的软件将包括一个框架和模块，使数据管理和可视化，信号处理和特征选择，分类和交叉验证。每个模块将代表一个计算工具，该工具已在第一阶段项目或第二阶段项目的研究阶段进行了评估，并根据数值稳健性、分类误差和处理时间等标准被选中。所开发的工具的模块化架构将确保软件包的可扩展性，并提供用于构建灵活的分类器的工具，这些分类器适用于数据采集协议、实验设备和蛋白质分析生物化学。我们的第一阶段研究表明，这种架构是最佳的研究和开发的分析MS论文在临床和学术环境，因为所需的灵活性和交互性水平。该软件包将授权给研究人员以及MS硬件制造商，以便与他们的仪器一起使用。

DESCRIPTION (provided by applicant): In the Phase II of this SBIR project we will develop a modular and extensible exploratory software tool for signal conditioning and data mining of profiling proteomics MS necessary for the development of cancer diagnostics. The software will allow researchers to construct and validate iterative classification methods for mass spectrometry (MS) data. The methods will be developed for surface enhanced laser desorption ionization (SELDI) MS, but the algorithms themselves and the modular framework will be applicable to analysis and classification of other types of MS data, such as matrix assisted laser desorption ionization (MALDI) and electrospray ionization (ESI). This classification software will be capable both of diagnosing disease and identifying clinically significant features in MS records. The user interface will be based on the existing INCOGEN VIBE software, which was developed to provide a user-friendly, extensible framework for data analysis. The proposed software will be comprised of a framework and modules that enable data management and visualization, signal processing and feature selection, and classification and cross-validation. Each module will represent a computational tool that has been evaluated in the Phase I project or during the research phase of the Phase II project and has been selected for inclusion based on criteria such as numerical robustness, classification error and processing time. The modular architecture of the developed tool will ensure extensibility of the software package and provide tools for construction of flexible classifiers that are adaptable to data acquisition protocols, experimental equipment and protein profiling bio-chemistry. Our Phase I research suggests that such architecture is optimal for research and development of profiling MS essays both in clinical and academic environment because of the required level of flexibility and interactivity. The software package will be licensed to researchers as well as to MS hardware manufacturers for inclusion with their instruments.