Pseudo-Reference Based Reconstruction of Avian and Primate Transcriptomes using High Throughput RNA Sequencing

High-throughput RNA sequencing (RNA-Seq) provides deep understanding of transcriptome in the view of the sort, structure, quantity, and biology of expressed transcripts in a cell. The short RNA-seq reads are assembled into transcripts by either reference-based or de-novo approach to annotate genes and isoforms. Although the reference-based approach guarantees a better quality of the resulting transcriptomes than does the de-novo approach, the reference-based approach is only applicable when the corresponding reference genome is available. Here, we developed the reference–based reconstruction method that maps reads to the genome of a closely related species with an optimized mapping parameter, named pseudo-reference based assembly. The optimal mapping parameters were estimated by the evolutionary distance models correlated to the mismatch rate of reads. Using the method, we reconstructed transcriptomes of four avian using Gallus gallus gallus genome and of three primates using human genome. To evaluate the quality of the resulting transcriptomes, we compared the resulting transcriptomes to the transcriptomes of the reference-based and de novo assemblies. We found that the pseudo-reference based assembly outperformed the de novoapproach within 0.363 evolutionary distance (calculated from D_loop region sequence) and was comparable to the reference-based approach within 0.418 evolutionary distance. Taken together, this approach can be widely applied in reconstructing transcriptomes of bacteria to mammals whose genomes are not sequenced. 

# This work was supported by a grant from Cooperative Research Program for Agriculture Science & Technology Development (Project title: National Agricultural Genome Program, Project No. PJ01045303).