GSK_2019

At Netrias, our team builds tools that make it easier for scientists and public health officials to work with data. A big challenge we see in the life sciences is that important datasets often come with inconsistent or confusing metadata related to how samples are described, how measurements are recorded, or how patient information is structured. These small differences can make it hard to combine or compare data across studies, which slows down progress.

We’d like to use the GSK AMR dataset as a real-world example to study how these inconsistencies show up in practice. We won’t train models on or share the data. Instead, we’ll analyze how the dataset is organized and use those lessons to improve our automated tools for cleaning up and standardizing data.

By doing this, we hope to:
1. Support better stewardship of AMR data by showing how harmonization can reduce wasted effort and errors.
2. Help public health practice by making it easier to integrate data from different studies and get a clearer picture of resistance trends.
3. Strengthen health systems by ensuring that data collected for one purpose can be reused more effectively for others.

Our goal is to make AMR data easier to use and more impactful, so it can directly support better science, stronger policy decisions, and, ultimately, better outcomes for patients.