SOURCE-LINKED INTELLIGENCE

Translational Control Decoded: Predictive, Interpretable and Generative Models of mRNA Regulation

CORDIS · observation · Publication date unknown

s mRNA features to correct translation levels. Yet, translation is characterized by a confined and modular cis-regulatory space, making it ideal for dissecting function with cutting edge genomics and machine learning (ML). TRANS-DECODE will deliver a systematic, data- and model-driven strategy for decoding the regulation of mRNA translation at unprecedented resolution and interpretability. Our innovative approach will combine high-throughput massively parallel reporter assays (MPRAs), ribosome profiling, and CRISPR based validation screens with explainable AI and RNA foundation models. We will use iterative design-build-test cycles in human cells and zebrafish as vertebrate model to construct interpretable, predictive models of translation initiation, elongation, and mRNA stability. The project is structured into four ambitious and synergistic aims: 1. Determine the RNA sequence/structure determinants of translation initiation in 5’UTRs using MPRAs and interpretable ML. 2. Connect sequence features to regulatory pathways by perturbing translation initiation factors. 3. Dissect spat

Read original source ↗ Open in workspace

recordType

award

status

SIGNED

region

EU

value

2487317

unit

EUR