Motif discovery within upstream regions of variable length reveals regulatory signatures in peach

Unraveling the mechanisms that regulate gene expression is a complex challenge in biology. A key step to decipher the regulatory machinery is to identify cis-regulatory elements (CREs) buried in non-coding DNA sequences. CREs are short stretches of DNA that serve as binding sites for transcription factors (TFs) and are frequently summarized as sequence motifs or logos. Although much attention has been paid to model species like Arabidopsis thaliana, little is known about regulatory motifs in other ones. Here, we describe a bottom-up approach for de novo motif discovery using peach as an example. These predictions require pre-computed gene clusters grouped by their expression similarity. Proximal promoter regions were defined as four different intervals: Up 1: [-1,500 bp, +200 bp], Up 2: [-500 bp, +200 bp], Up 3: [-500 bp, 0 bp] and Up 4 [0 bp, +200 bp]. Two algorithms from RSAT::Plants (http://plants.rsat.eu) were tested (oligo and dyad analysis). Overall, 18 out of 45 co-expressed modules were enriched in motifs typical of well-known TF families (bHLH, bZip, BZR, CAMTA, DOF, E2FE, AP2-ERF, Myb-like, NAC, TCP, and WRKY) and a few uncharacterized motifs. Our results indicate that small module size and promoter window of [-500 bp, +200 bp] relative to the transcription start site (TSS) maximize the number of motifs found and reduce low-complexity signals in peach. This work yields a comprehensive collection of Prunus persica motifs without prior knowledge and provides a pipeline that can be applied to other species.