hide
Free keywords:
-
Abstract:
Heterosis refers to the deviation of the F1 progeny from the phenotypic mean of the parental plants, especially in cases where both parents are inbred and homozygous throughout the genome. When comparing certain traits of a F1 hybrid to the corresponding parental plants one can compare the performance of the hybrid to (1) the midparent value (MPV), which refers to the mean of both parental plants for the given trait. Furthermore the F1 hybrid can be compared to (2) the best-parent value (BPV), referring to the value of the superior parental plant. Mainly three genetic models, explaining heterosis have been proposed. However, there is no consensus about the diversity of molecular principles of heterosis. Recent studies suggested structural genome variation, such as copy-number variation (CNV) as well as presence absence variation (PAV), among maize inbred lines, underlying complementary contributions of genes from both parents as an important factor (Springer et al. 2009). Gene expression with regard to heterosis has been analyzed in several plants includingA. thaliana (Alonso-Peral et al. 2017). However, the majority of these studies relied either on microarray data, which are limited to transcripts resent on the chip, or on RNA-seq data that have been processed using a single reference genome, thus not accounting for structural variation among different genotypes (accessions). We now have access to high quality full length genome sequences of different A. thaliana accessions. Two of these accessions have been used in a reciprocal crossing. The root and shoot transcriptomes of F1 hybrids as well as from the corresponding parental plants have been short read sequenced. RNA-seq reads have been processed using either one parental genome as a reference or a trans-reference genome, containing full length genome sequences of both parental plants and accounting for orthologous assignment. DESeq2 has been used to detect transcripts with non parental expression patterns (below or above MPV) in the F1 hybrid plants. When using a single reference genome we found various transposable elements among the differentially expressed genes (p < 0.01). However, whole genome alignment data among parental plants indicate that many of these are due to a reference bias. Here we present a way of how to process a double reference genome in order to improve the analysis of hybrid transcriptomes regarding heterosis.
