A comprehensive annotation and differential expression analysis of short and long non-coding RNAs in 16 bat genomes

Nelly Mostajo Berrospi; Marie Lataretu; Sebastian Krautwurst; Florian Mock; Daniel Desirò; Kevin Lamkiewicz; Maximilian Collatz; Andreas Schoen; Friedemann Weber; Manja Marz; Martin Hölzer

In Tab. S5 we list all detected snoRNAs, divided into box C/D and box H/ACA types. Overall, we found 162 snoRNA families within the investigated bat species, comprising 88 box C/D, 61 box H/ACA, and 13 unclassified snoRNAs. Many snoRNAs were found with exactly one copy present in each bat genome assembly (e.g. SCARNA7), whereas others were found in multiple copies for each bat species (e.g. SNORD112) or completely absent from certain bat families (e.g. ACA64), see Tab. 5. Exactly one copy of the small nucleolar RNA ACA64 was found within the genomes of the Pteropodidae family and multiple copies for D. rotundus, P. parnellii, M. natalensis and members of the Vespertilionidae family, however, this snoRNA seems to be completely absent from bat species of the Megadermatidae and Rhinolophidae families (Tab. 5). The H/ACA box snoRNA ACA64 is predicted to guide the pseudo-uridylation of 28S rRNA U4331 63 . Interestingly and as another example, SNORA48 was found in higher copy numbers in the genomes of the Vespertilionidae family. Among others, this H/ACA box snoRNA was described to be commonly altered in human disease 64 . The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/738526 doi: bioRxiv preprint 400 miRNAs were predicted. For the other 4 species of the Vespertilionidae family about 5 times as many miRNAs could be found. This is in concordance with the small RNA-Seq data used for the prediction, that was obtained from Myotis daubentonii kidney cells (see Methods). Nevertheless, ∼400 conserved miRNAs can be predicted in the evolutionary more distant bat species. We compared our miRDeep2-predictions in M. lucifugus and P. alecto with the predictions of two other studies 8, 55 . lncRNAs For the annotation of lncRNAs, we have deliberately chosen a broad Blast-based approach, using 107,039 transcripts of potential human lncRNAs obtained from the LNCipedia data base 56 . We have consciously chosen this approach, because lncRNAs have diverse genomic contexts, reveal various functions and act in different biological mechanisms [65] [66] [67] .