Figure 6. Pipeline for detection of pRNAs. pRNAs are searched by homology and aligned. New candidates (passing the conservation and synteny filter) are used additionally and iteratively as input for the next search step (black). For each next search iteration a MEME-derived motif for A-box and B-box is used independently (green). We repeated these steps until no new trustable sequences were obtained. In the end (blue), the promising candidates, homologous sequences from NCBI databases EST, WGS or HTGS, as well as described pRNAs from the literature are taken together and copies are removed. The final pRNA-set is presented in the last column of Table 1 and in detail in the Supplemental Material. Parameters are given in main text.