A computational framework to study single-molecule canonical and non-canonical translation dynamics.

Advances in fluorescence microscopy allow the direct visualization of gene expression with single-molecule resolution in live cells. Here, we combine fluorescence microscopy and stochastic modeling to study canonical and non-canonical ribosomal translation processes. Recent findings by our groups allowed us to conduct an extensive computational study of mRNA-translation at single-codon resolution in a large database of human genes. We integrated the results from this study in a software package rSNAPsim (Aguilera, et al., 2019, Plos Comp Biol). In a second study, we quantified ribosomal frameshifting in live cells at single-molecule resolution. Experimental data and simulations proved that frameshifting is a rare process occurring in ~8% of the studied proteins, occurring in long-lasting (~40min) bursting episodes (Lyon, et al., 2019. Mol Cell). More recently, our groups used a bicistronic biosensor to measure Cap-dependent and IRES-mediated initiation. We obtained that under normal conditions CAP initiation is ~ 3 times stronger than IRES initiation, but this proportion is dramatically altered during stress conditions that mimic viral infection (Koch, et al., Nat Struct Mol Biol 2020).