Revolutionizing Biomedical Research with Federated Machine Learning

Data Privacy Challenges in Biomedical Research

One of the largest challenges in the age of -omics is the privacy and sharing of data. Robust statistical analyses and machine learning models benefit hugely from using the largest sample sizes possible. However, most research entities are grounded in the principles of maintaining control and ownership of their own datasets.

This poses a particular challenge for studying rare diseases, where datasets are few and far between, and without sharing data from multiple centers, valuable disease insights may be lost due to a lack of sufficient power. A novel machine learning paradigm called collaborative federated learning provides a unique solution: a secure software is installed in each center that allows machine learning models to be trained by seeing the data from each center without physically pooling the data in a central repository. 

This federated learning approach maintains individual control and ownership of data while allowing machine learning models to indirectly access the data to strengthen their predictive power and accuracy. An approach like this has the potential to revolutionize biomedical research, particularly for rare diseases, by making larger datasets available for model training.

Applying Federated Learning to Triple Negative Breast Cancer Biomarkers

In a recent paper in Nature Medicine, Ogier du Terrail et. al. performed a proof-of-concept study for federated learning to build a predictive model for triple negative breast cancer response to neoadjuvant chemotherapy (NACT). While NACT is the current standard of care for this rare, aggressive type of breast cancer, there is significant variety in patient response which is still poorly understood.

When trained on the collaborated, federated learning data, the ML model developed by the authors predicted breast cancer response to NACT nearly as well as the current best models developed using meticulous expert annotation. This approach allows for the interrogation of data in unprecedentedly large quantities, hopefully revealing patterns of disease and robust new biomarkers.

Outsourcing Bioinformatics Analysis: How Bridge Informatics Can Help

Our clients are tackling their research questions using sophisticated machine learning approaches in bioinformatics, and many more cutting edge tools. As experts across data types from cutting-edge sequencing platforms, we can help you tackle the challenging computational tasks of storing, analyzing and interpreting genomic and transcriptomic data. Bridge Informatics’ bioinformaticians are trained bench biologists, so they understand the biological questions driving your computational analysis. From pipeline development and software engineering to deploying existing bioinformatics tools, Bridge Informatics can help you on every step of your research journey. Click here to schedule a free introductory call with a member of our team.



Jane Cook, Biochemist & Content Writer, Bridge Informatics

Jane Cook, leading Content Writer for Bridge Informatics, has written over 100 articles on the latest topics and trends for the bioinformatics community. Jane’s broad and deep interdisciplinary molecular biology experience spans developing biochemistry assays to genomics. Prior to joining Bridge, Jane held research assistant roles in biochemistry research labs across a variety of therapeutic areas. While obtaining her B.A. in Biochemistry from Trinity College in Dublin, Ireland, Jane also studied journalism at New York University’s Arthur L. Carter Journalism Institute. As a native Texan, she embraces any challenge that comes her way. Jane hails from Dallas but returns to Ireland any and every chance she gets. If you’re interested in reaching out, please email jennifer.martinez@old.bridgeinformatics.com or dan.ryder@old.bridgeinformatics.com.

Recent Posts