Quality control assessment of the RNA-Seq data generated from liver and pituitary transcriptome of Hereford bulls using StrandNGS software

Abstract

Background: Quality control (QC) assessment is the most critical step in the high-throughput RNA-seq data analysis to characterize the in-depth understanding of genome and transcriptome assembling to a given reference genome. It provides not only a quick insight into the RNA-seq data quality to allow early identification of good or bad RNA-seq data samples, but also to verify the alignment QC checks for further essential high-throughput bioinformatics analysis such as, identification of novel genetic variants, differentially expressed genes (DEGs), gene network and metabolic pathways.

Method: After isolation of total RNA from liver (n=15) and pituitary gland (n=15) tissues of young Hereford bulls, the pooled total RNA (n=30) were fragmented using GeneRead rRNA depletion kit (Qiagen, Hilden, Germany) and cDNA library preparation were preformed using ScriptSeq^TM v2 RNA-Seq library preparation kit (Epicentre, illumina, USA), followed by high-throughput sequencing of combined liver and pituitary transcriptome using MiSeq reagent kit v2 (illumina, USA) to obtain high quality of paired-end RNA-seq reads of 251 base-pairs (bps). In this paper, the QC assessment of obtained RNA-seq raw data as well as post-alignment QC of processed RNA-seq data of combined liver and pituitary transcriptome (n=30) of Hereford bulls were performed using the strand NGS software v1.3 (Agilent; http://www.strand-ngs.com/) data analysis package. The reads were aligned with Bowtie using default settings against both Bull and Cow genome assembly.

Results: Using two runs of MiSeq platform, a total of over 60 million paired-end RNA-seq reads were successfully obtained and submitted to NCBI SRA resources (https://www.ncbi.nlm.nih.gov/sra?linkname=bioproject_sra_all&from_uid=312148). Library complexity plot results revealed 72.02% of duplicate reads with a low library complexity value of 0.28. The pre-alignment QC analysis of raw RNA-seq data revealed the sequence read lengths ranged from 35-251 bp size with more than 50% of all reads with length over 200bp and 10% of reads below 100bp.

Conclusion: By testing the RNA-seq methodology on Illumina platform, two MiSeq sequencing runs yielded significantly high quality of 30 million sequencing reads per single MiSeq run. Our initial pre-alignment and post-alignment analysis of RNA-seq data analysis revealed that mapping of the Hereford liver and pituitary gland transcriptome to reference Bos taurus genome was successfully performed, however, more than 50% of all reads with length over 200bp were recovered. Therefore, obtained results concludes that liver and pituitary transcriptome sequencing with rRNA depletion method is less effective than mRNA RNA-seq method.