Yana Bromberg

Career and research

Bromberg started her academic career as an Assistant Professor of Bioinformatics at the Rutgers School of Environmental and Biological Sciences (2010). She finally joined the faculty at Emory University as a Full Professor in 2023. She holds joint appointments in the departments of Biology and Computer Science and serves as a principal fellow at the Center for AI Learning.

Bromberg's research primarily focuses on building machine learning and artificial intelligence -based methods for understanding of the origins of life, annotation of the microbiome functionality, genetic variant impact and protein function prediction. Her lab develops novel bioinformatics techniques integrating machine learning, high-performance computing, and large-scale genomic and metagenomic analysis, with particular emphasis on protein molecular functions.

SNAP (published in 2007), developed as part of her doctoral work, was one of the first tools to use neural networks for the analysis of genomic variants (here, non-synonymous SNPs). Her lab had since developed SynVep, a machine learning-based predictor for evaluating the impact of synonymous variants. Recently, the Bromberglab had demonstrated that unsupervised pLMs can be used to predict variant effects as well as earlier, supervised methods, albeit no existing model attains perfect understanding of variant impact.

A central theme of her recent work is investigating what biological information is captured by pLMs. To this end, her lab has developed the RNS method to quantify the uncertainty in protein representations, indicative of their ability to genuinely encode protein biology.

Selected talks

• "The Big YOU: Defining 'you' on the Microbial Level" - TEDxRutgers (2019)
• "Yana Bromberg on getting creative with machine learning" - Night Science Podcast (June 2021)
• Function SIG presentation - ISMB/ECCB 2023, Lyon, France (July 2023)
• 3DSIG presentation - ISMB 2024, Montreal, Canada (July 2024)