650 / 2019-04-29 18:46:53

Using PacBio Isoform Sequencing data to improve the de novo assembly of RNA-seq data

PcaBio,RNA-seq,Transcriptome

全体主题 > Omics& New Technology

De novo assembly of RNA-seq data provides a means to study transcriptomes without a reference genome sequence. However, assembling short reads is highly challenging because of the complexity of the transcriptome. Therefore, we sequenced the leaf transcriptome of Gynandropsis gynandra, a species related to the model plant Arabidopsis thaliana, by PcaBio Iso-seq and Illumina RNA-seq and established a pipeline to improve the short-reads de novo assembly. First, the longest transcripts were selected as representative transcripts from PacBio data. Additionally, the CLC Genomics Workbench was used to do RNA-seq assembly. Then, we predicted and collected the PacBio and CLC CDSs each of which covers over 90% of the length of its targeting Arabidopsis gene into a CDS dataset, calling it as a full length CDS dataset. Second, the unmapped RNA-seq reads against the full length CDSs were used to do de novo assembly. After assembling, new full length CDSs were collected into the full length CDS dataset. Third, the remaining transcripts from PacBio Iso-seq and CLC contigs were used to assemble by CAP3. Finally, we constructed a new G. gynandra CDS dataset in which we combined CAP3 results with the full length CDS dataset and removed redundant transcripts. After these steps, we reconstructed over 80% of CDSs at full length in G. gynandra leaf transcriptome, which significantly improves the de novo assembly result compared to the assembly using only the RNA-seq data by current assembly tools.