tRNAviz: explore and visualize tRNA sequence features

Abstract

Transfer RNAs (tRNAs) are ubiquitous across the tree of life. Although tRNA structure is highly conserved, there is still significant variation in sequence features between clades, isotypes and even isodecoders. This variation not only impacts translation, but as shown by a variety of recent studies, nontranslation-associated functions are also sensitive to small changes in tRNA sequence. Despite the rapidly growing number of sequenced genomes, there is a lack of tools for both small- and large-scale comparative genomics analysis of tRNA sequence features. Here, we have integrated over 150 000 tRNAs spanning all domains of life into tRNAviz, a web application for exploring and visualizing tRNA sequence features. tRNAviz implements a framework for determining consensus sequence features and can generate sequence feature distributions by isotypes, clades and anticodons, among other tRNA properties such as score. All visualizations are interactive and exportable. The web server is publicly available at http://trna.ucsc.edu/tRNAviz/.

Figure 1.

Consensus features are displayed on the Summary page. (A) A tRNA cloverleaf is visualized with the consensus sequence features of tRNA^Ala in primates. (B) A tilemap displays consensus isotype-specific features for primates. On the web server, mouseover of tiles (or cloverleaf positions) dynamically provides the base frequency distribution for the corresponding isotype and position. Colors correspond to nucleotide color legend.

Figure 2.

Compare tRNAs across species for focused isotypes/positions. Sequence feature distribution comparison shown between two clade groups (i) genera Saccharomyces and Schizosaccharomyces, and (ii) Basidiomycota phylum, at base pair 3:70 in tRNA^Gly and tRNA^Asn. Bars represent a stacked histogram with colors corresponding to different nucleotide features. Saccharomyces have G:C or C:G exclusively at position 3:70 for these isotypes, whereas Schizosaccharomyces only use A:U or U:A at these positions. Basidiomycota species, by contrast, have a highly varied mix of base pair types at these positions. On the web server, mouseover of distribution bars reveals more detailed information about the specific feature.

Figure 3.

Bitchart and sequence feature distribution comparison for tRNA^Cys(GCA). (A) Position-specific sequence features with corresponding scoring scale for Cys-GCA-19 and Cys-GCA-24 in human are displayed in comparison with the consensus features in all human, mouse, primates and mammalian tRNA^Cys(GCA). The color scale of the features represents the conservation level with the reference. (B) Sequence feature distribution of tRNA^Cys(GCA) for six positions among five different clades is shown for comparison. The rows in the histogram represent the different positions of tRNA^Cys(GCA). The vertical bars (from left to right) within each row represent the five clades listed in the clade group table. The colors in the histogram represent the frequency of the corresponding sequence features (listed in the legend) at a particular position of a clade.