“…This hypothesis is based on the fact that hundreds of smORFs are described as highly conserved but display low expression, low translation efficiency and are observed in transcripts with non-coding characteristics (Cabili et al, 2011;Aspden et al, 2014;Bazzini et al, 2014). However, the nearly neutral theory (Ohta, 2002) suggests that non-coding parts of fixed smORF transcripts are modified by random genetic drift, in some cases, producing small advantageous (or disadvantageous) adaptive effects throughout evolution; thus, we propose that, at a certain point, these modifications refine and elevate the coding potential of smORF transcripts and consequently enhance the adaptive relevance of their peptides, as seen in a large number of important smORF peptides recently discovered (e.g., Magny et al, 2013;Anderson et al, 2015;Lauressergues et al, 2015;Nelson et al, 2016;Pengpeng et al, 2017;Kim et al, 2018;Polycarpou-Schwarz et al, 2018;Chugunova et al, 2019;Tobias-Santos et al, 2019;Pang et al, 2020;Vassallo et al, 2020). Importantly, the acquisition of several optimal coding features might be favored after the smORF has been selected for, because modifications driven by genetic drift could be fixed by natural selection if they improve the translation efficiency of the newly selected smORF.…”