Biostatistics Core

摘要：生物统计学核心 生物统计学核心（BC）将支持研究设计、数据处理和个体分析需求。 项目，并加强P20中所有项目之间的互动。经验表明 生物统计学家和数据科学家从概念阶段的参与产生了更好的研究 设计，更有可能回答感兴趣的科学问题，并最终，在他们的 结论.生物统计学核心的目标是在研究的所有阶段为研究者提供支持 从设计到出版，通过以下具体目标。 具体目标1：研究设计 BC将采用5阶段早期检测生物标志物开发指南以及PRoBE研究 设计标准，以帮助P20团队建立路线图和研究战略， AI/AN患者的肝癌早期检测。5阶段指南的使用将允许P20研究者 制定战略并制定长期计划，帮助生物标志物从发现阶段1 - 2进入验证阶段 3 - 5，实现平移连续体。遵守PRoBE标准将指导研究设计， 预期的临床应用，减少系统偏倚，并确保重现性和 临床相关研究结论。 具体目标2：分析和解释 BC将在拟议的Li-CAD P20项目的统计分析和解释方面发挥领导作用。 巴塞尔公约将确定和实施量化方法，以解决感兴趣的科学问题， 提供关于解决各种研究假设的证据的有效统计推断。比如说， 将使用参数经验贝叶斯（PEG）和多变量全贝叶斯（mFB）方法进行建模 纵向生物标志物，将用于建立AI/AN中HCC早期检测的风险预测模型 人口此外，将进行数据可视化以研究HBV突变， 机器学习方法将用于开发和验证人工智能/人工智能中的HCC风险预测模型 人士BC还将与研究者合作，明确传达研究方法和结果 出版物，并确保报告的结论是合理的。 具体目标3：促进与DRP和其他核心的互动 BC将与DRP以及生物标本和病理学核心互动，以确保 Li-CAD P20正在进行中，包括所有拟议的P20项目以及DRP项目。BC领导人将 直接与P20执行委员会沟通，以确保任何有统计影响的决定 从核心得到适当和及时的投入。

ABSTRACT: BIOSTATISTICS CORE The Biostatistics Core (BC) will support the study design, data processing, and analysis needs for individual projects as well as enhance interactions among all the projects in the P20. Experience has shown that involvement of biostatisticians and data scientists from the concept phase yields studies that are better designed, more likely to answer the scientific questions of interest, and, ultimately, more compelling in their conclusions. The goal of the Biostatistics Core is to support the investigators through all phases of their studies from design to publication via the following Specific Aims. Specific Aim 1: Study Design The BC will employ the 5-phase early detection biomarker development guideline as well as the PRoBE study design standards in order to help the P20 team establish a road map and research strategies for improving liver cancer early detection in AI/AN patients. The use of the 5-phase guideline will allow P20 investigators to strategize and make long-term plans, helping move biomarkers from discovery phases 1-2 to validation phases 3-5, fulfilling the translational continuum. Adhering to PRoBE standards will guide the study design driven by intended clinical application, reducing systematic bias, and ensuring a high likelihood of reproducible and clinically relevant study conclusions. Specific Aim 2: Analysis and Interpretation The BC will play a leadership role in statistical analysis and interpretation for the proposed Li-CAD P20 projects. The BC will identify and implement quantitative methods to address the scientific questions of interest and provide valid statistical inferences about the evidence addressing the various study hypotheses. For example, Parametric Empirical Bayesian (PEG) and Multivariate Fully Bayesian (mFB) approaches will be used to model longitudinal biomarkers, which will be used to build risk prediction models for the early detection of HCC in AI/AN population. Additionally, data visualization will be conducted to investigate HBV mutations, and state-of-the-art machine learning approaches will be utilized to develop and validate HCC risk prediction models in AI/AN persons. The BC will also work with study investigators to clearly communicate methods and results in study publications and insure that reported conclusions are justified. Specific Aim 3: Promote Interactions with DRP and Other Cores The BC will interact with the DRP as well as the Biospecimen and Pathology Core to ensure efforts within the Li-CAD P20 are on track, including all proposed P20 projects as well as DRP projects. The BC leader will communicate directly with the P20 Executive Committee to ensure any decision that has statistical implications has appropriate and timely inputs from the Core.