Trends of disparities in breast cancer progression and health care considering multilevel risk factors

点击翻译按钮获取中文摘要

A Scalable Cloud-based Platform for Multiple Mediation Analysis The primary objective of the parent award (#MD012387-02) is to examine the trend of racial disparity in the use of Oncotype diagnosis among breast cancer patients. It also aims to monitor how social and physical environments, as well as individual behaviors, contribute to the racial disparities in breast cancer outcomes. To achieve these goals, the parent award focuses on developing novel mediation and moderation analysis methods using state-of-the-art machine learning algorithms and software. These methods can help identify and differentiate risk factors that explain health disparities and their changes over time, considering the potential nonlinear associations and complicated interaction effects involved. As part of the parent award, an R package mma has been generated for the newly developed method. The goal of this administrative supplement proposal is to expand the use of mma by collaborating with a software engineer in two ways. Firstly, build an interactive visual interface with drag-and-drop functionalities to enable users who are not familiar with the R language to take advantage of the R package. Secondly, refactor the mma functions to enable the use of parallel computing and GPU support in the cloud high-performance computing environment. In summary, the proposal seeks to make the advanced technology accessible and promote the widespread adoption of the novel mediation and moderation method among researchers.