This is a key step in delivering on the promise of using AI for designing more sustainable crops! Big thanks to all the co-authors, especially leads Nicolas Kral and Sania Jevtic!

2. We find many promoters with overexpression and validate top candidates at the protein level! And 3. without teaching the model anything about transcription factor motifs, we see that it learns motif syntax!

1. We introduce a new AI model for predicting plant gene expression with competitive performance. We then use the model to evolve and test thousands of promoters for over-expressing a drought-resilience trait in tomato protoplasts using an MPRA 🍅