Variant calling from RNA-seq data of the brain transcriptome of pigs and its application for allele-specific expression and imprinting analysis

Abstract

Identification of new polymorphic variants from RNA-seq data is difficult mainly because of the errors arising during bioinformatic analysis. Therefore, new experiments in this area are very profitable for improving new statistical methods. In our study of the porcine brain transcriptome, we have identified 10966 polymorphic variants, among which 7277 were single nucleotide polymorphisms (SNPs). Further, we have calculated allelic ratios for the SNPs identified and estimated that 52% of genes in porcine brain are subjected to allele-specific expression (ASE), a phenomenon in which one allele is preferentially expressed. Our investigation presents the first estimates of ASE in porcine brain. In addition, we have used the results of RNA-seq for the identification of SNPs in putatively imprinted genes. Finally, we have used these SNPs for the verification of the imprinted status of the INPP5f variant 2, LRRTM1 and HM13 genes in pigs by Sanger sequencing. We observed that INPP5f variant 2 is paternally expressed, while HM13 and LRRTM1 are biallelically expressed in porcine brain. We have also confirmed maternal expression of the MEG3 gene in pigs. Our results present how RNA-seq data may be used for imprinting studies without sequencing of parental genomes.

Keywords: Allele-specific expression; Imprinting; Pigs; RNA-seq; SNPs; Variant calling.