Figure 2

Properties of rice lncRNAs. (A) The proportion of lncNAT sequences overlapped by mRNAs. (B) The number of exons per transcript for all lincRNAs and lncNATs and protein-coding transcripts. (C) Exon size distributions for lincRNAs, lncNATs and protein-coding transcripts. (D)Transcript size distributions for lincRNAs, lncNATs and protein-coding transcripts. (E) The fraction of transcripts overlapping at least one base of a repetitive element (left), and the fraction of all transcript sequences overlapping repeats (right). The repeat elements found in rice were annotated using RepeatMasker [28], whereas the repeat elements found in human sequences were downloaded from the UCSC genome browser [29] and were also annotated by RepeatMasker. (F) A/U content of rice lincRNAs and lncNATs and various regions of protein-coding transcripts. (G) Correlation between the expression levels of lincRNAs and the expression levels of their closest protein-coding genes. The Spearman correlation between the expression levels of each gene across 17 stages/tissues and the expression levels of the closest protein-coding gene were calculated. Average values are shown in the plot. Error bars represent standard deviations based on 1,000 random shuffles of lincRNA positions. (H) The fraction of residues aligned in multiple-genome alignments (MGA) for the indicated protein-coding gene, TE-mRNA and lincRNA regions. The controls are random intergenic regions that were size- and-chromosome matched to the lincRNA set.CDS, coding sequence. (I) The level of conservation of the exons, introns and UTRs (if any) of protein-coding genes, TE-mRNAs and lincRNAs. The cumulative distributions of mean phastCons scores derived from eight-way whole-genome alignments are shown.