P019
L_RNA_scaffolder: Scaffolding Genomes with Transcripts
L_RNA_scaffolder: Scaffolding Genomes with Transcripts
We describe L_RNA_scaffolder, a de novo genome scaffolding method that uses long trans-criptome reads to order and orient genomic fragments and combine them into larger sequences. When one transcript was not fully covered by a genomic fragment, adjacent exons anchored two genomic fragments and this was then used as evidence of a linkage between the fragments. This process is similar to how pair-end reads are used in genome scaffolding. From the Ensembl annotations of five well assembled genomes, 4.3%–26.4% of the introns are estimated to be longer than 5kb. Although this proportion is low, these large introns cover 42.5%–83.9% of the transcribed genomic regions. Further, mounting evidence has suggested that the vast majority of the genome is transcribed. Therefore, transcripts from the pervasively transcribed genomic regions might function as long-insert libraries to scaffold most of the transcribed genome regions. The zebrafish genome was scaffolded to demonstrate the accuracy of the method. After scaffolding with the expanded human transcriptome data, the N50 was doubled. In these two examples, the transcript coverage was both improved and close to the complete genomes. L_RNA_scaffolder performance was better than most large-insert libraries. The simplicity and high-throughput data of RNA-s eq technology make this approach suitable for genome scaffolding. L_RNA_scaffolder pipeline is available at http://www.fishbrowser.org/software/L_RNA_scaffolder.
