High-throughput interrogation of programmed ribosomal frameshifting in human cells

Martin Mikl; Yitzhak Pilpel; Eran Segal

In general, our results reveal functional groups of PRF events that exhibit greater similarity in terms 577 of their sensitivities to mutations in regulatory regions and congruence in the preferences for 578 secondary structure. These functional groups also manifest themselves in a higher prediction 579 accuracy within these groups. The groups resemble similarities also according to other criteria, e.g. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It . https://doi.org/10.1101/469692 doi: bioRxiv preprint higher fluorescence signal in our assay (Fig 1C) , but greater sensitivities to even small sequence 586 changes (Fig 2B, 3C, S4AB, S5A ) and requirements for a more extensive downstream region. This 587 might also entail a greater sensitivity to the experimental context and therefore explain the lack of 588 signal (above the noise threshold) for previously reported cases depending on a pseudoknot as 589 frameshift stimulatory signal like Rous sarcoma virus (Jacks and Varmus, 1985; Marczinke et al., 590 1998) . In other cases, low wild-type rates of frameshifting, requirements for additional trans-acting 591 factors or an effect on overall expression levels might explain the lack of signal in our assay. 592