Cancer Informatics Shared Resource

项目总结/摘要 癌症信息共享资源（CISR）的使命是促进高质量，创新的癌症 研究通过为威斯康星州大学Carbone癌症中心（UWCCC）的成员提供切割- 边缘和高效信息学，包括生物信息学，临床信息学，图像分析，机器学习， 和计算科学。CISR通过一个由五名高技能博士生组成的团队完成这一使命 科学家，四个世界著名的信息学教师，以及大量的计算系统和支持， 威斯康星州大学生物统计学和医学信息学系（BMI），其中CISR是 成立于2003年。具体目标1。提供最先进的信息服务。CISR提供了深入和 在生物信息学和生物、临床和成像数据的机器学习分析方面拥有广泛的专业知识。CISR 工作人员精通各种任务的最新计算方法， 实施和应用它们来支持UWCCC成员项目。这种支持是通过专门的 CISR工作人员对成员资助项目的努力; CISR工作人员参与研究设计，试点研究， 赠款申请开发;以及开发标准化计算管道。具体目标2。 开发创新的信息学方法来支持癌症研究。CISR工作人员开发新的 计算方法和算法，使UWCCC的研究，以达到其全部潜力时，现有的 方法还不够。支持这一目标是由四名教师的努力与专业知识的支持 在机器学习，电子健康记录分析， 从观察数据、自然语言处理、深度学习和RNA-Seq基因中发现因果关系 表达数据分析。教师都与合作，并指导博士科学家的发展， 新的计算方法及其在成员研究中的应用。在目前的癌症中心 支助赠款周期，随着工作人员从3人增加到5人，对这一目标的支助得到加强 科学家，进一步提高能力和扩大CISR的专业知识，包括免疫信号， 基因组注释、转录调控网络和表观遗传学。在本周期内，CISR支持了61个 所有六个科学计划的UWCCC成员和使用量比前一个周期翻了一番， 平均每年服务的成员人数。与UWCCC调查人员，CISR工作人员共同撰写了34 癌症相关的文章，有助于提交30个赠款申请，并支持赠款总额 1400万美元的直接成本，其中5个来自NCI，4个来自其他NIH研究所。CISR发挥了重要作用， 在几个大规模的UWCCC项目中，包括头颈癌孢子（P50 DE 026787）和 靶向放射性核素治疗P01（P01 CA 250972）。CCSG对CISR的资助将确保其团队 继续处于有利地位，以满足UWCCC研究人员不断发展的计算研究需求， 尖端的癌症研究，这是越来越依赖于大规模和异构的数据集。

PROJECT SUMMARY / ABSTRACT The mission of the Cancer Informatics Shared Resource (CISR) is to promote high quality, innovative cancer research by providing members of the University of Wisconsin Carbone Cancer Center (UWCCC) with cutting- edge and efficient informatics, including bioinformatics, clinical informatics, image analysis, machine learning, and computational science. CISR accomplishes this mission through a team of five highly skilled PhD staff scientists, four world-renowned informatics faculty, and the substantial computational systems and support of the University of Wisconsin Department of Biostatistics and Medical Informatics (BMI), within which CISR was established in 2003. Specific Aim 1. Provide State-of-the-art Informatics Services. CISR provides deep and broad expertise in bioinformatics and machine learning analysis of biological, clinical, and imaging data. CISR staff are well versed in the latest computational approaches for a wide variety of tasks and can efficiently implement and apply them to support UWCCC member projects. This support is provided through dedicated effort of CISR staff to members’ funded projects; involvement of CISR staff in study design, pilot studies, and grant application development; and development of standardized computational pipelines. Specific Aim 2. Develop Innovative Informatics Methodology to Support Cancer Research. CISR staff develop novel computational approaches and algorithms to allow UWCCC research to reach its full potential when existing approaches are insufficient. Support for this aim is bolstered by the effort of four faculty members with expertise in the development of novel algorithms in areas such as machine learning, electronic health record analysis, causal discovery from observational data, natural language processing, deep learning, and RNA-Seq gene expression data analysis. Faculty both collaborate with and mentor the PhD scientists in the development of novel computational approaches and their application to members’ research. During the current Cancer Center Support Grant (CCSG) cycle, support for this aim was enhanced with the growth of the staff from three to five scientists, further increasing capacity and expanding the expertise of CISR to include immunological signaling, genome annotation, transcriptional regulatory networks, and epigenetics. During this cycle, CISR supported 61 individual UWCCC members across all six scientific programs and usage doubled over the prior cycle, in terms of the average number of members served per year. With UWCCC investigators, CISR staff co-authored 34 cancer-related articles, contributed to the submission of 30 grant applications, and supported grants totaling $14M in direct costs, including five from the NCI and four from other NIH institutes. CISR has had important roles in several large-scale UWCCC projects, including the Head and Neck Cancer SPORE (P50DE026787) and a P01 for targeted radionuclide therapy (P01CA250972). CCSG funding of CISR will ensure that its team will continue to be well positioned to meet the evolving computational research needs of UWCCC investigators for cutting-edge cancer research, which is increasingly reliant on massive and heterogeneous data sets.