Data Science Core

点击翻译按钮获取中文摘要

Biomedical research has experienced a crisis of reproducibility, exacerbated by lack of coordination between research groups, weak study designs, opaque algorithms, and an unwillingness to share the fruits of research efforts. But recent years have seen the beginning of a push to solve this crisis, with a focus on integration across labs, standardization, rigorous statistical analysis, open source code, and the sharing of data and methods. This proposal aims to “Crack the Olfactory Code”, a far-reaching goal that will require a firm commitment to rigorous, open science principles. The Data Science Core (DSC) for this proposal reflects a plan to meet this commitment in three major ways: (1) The DSC will tightly coordinate the integration of data across the proposal’s five scientific projects, utilizing standards and adopting technology for total reproducibility of all findings, as well as ensuring that all research outputs will be made comprehensively and intelligibly available to the public; (2) The DSC will work with project leaders to optimize data collection decisions using statistical models, saving time and money while increasing scientific inference; (3) The DSC will standardize odorant stimuli across projects and create a service for pushing this standard into the larger research community, ensuring that future research efforts can be directly informed by our efforts and can inform each other.