Citation: M. Haas (2018-12-12): Linear modeling reveals a predominance of cis- over trans-regulatory effects in wild and domesticated barley. DOI:10.5447/ipk/2020/1

Abstract: Barley, like other crops, has experienced a series of genetic changes that have impacted its architecture and growth habit to suit the needs of humans. Domestication also resulted in a concomitant bottleneck that reduced sequence diversity in genes and regulatory regions. Little is known about regulatory changed resulting from domestication in barley. We used RNA-seq to examine allele-specific expression (ASE) in hybrids between wild and domesticated barley. Our results show that most genes have conserved regulation. In contrast to studies of ASE in interspecific hybrids, we find almost a complete absence of trans effects. We also find that cis regulation is largely stable in response to short-term cold stress. Our study has practical implications for crop improvement using wild relatives. Our experimental design allows us to compare the genome-wide expression of two individuals to each other as well as the allele-specific expression of the two alleles of each parent in a single hybrid, allowing for the precise measurement of cis regulatory effects. The accessions used in this study vary by domestication status (4 wild accessions, 2 landraces, and 2 European cultivars). All crosses were made between one of these accessions and the six-rowed American malting cultivar Morex. Morex is also the cultivar which was used to generate the first barley reference genome, which facilitated read-mapping. Exome capture and/or whole-genome shotgun data were used to call informative single nucleotide polymorphisms (SNPs) using standard variant callers. These SNPs allowed us to unambiguously assign reads to either allele. A three-hour cold treatment (4 degrees Celsius) was used to induce a stress response and search for differentially expressed genes in response to cold. There are four replicates for most genotype x generation x treatment combinations, although for some only 2-3 replicates are available. For one combination (Barke hybrid cold), only a single sample is available: after multiple attemps, the other replicates were not successful. The data included here are: 1) the expression matrix (181203_expression_matrix.csv), 2) gene information (181203_gene_info.csv), and 3) information about each sample subjected to RNA-seq (181203_sample_info.csv). There are also 8 subdirectories containing results from the limma analysis. Each directory name has a prefix with the name of the individual accession used for the analysis. Each subdirectory contains 16 csv files: 1) accession..useless info, 2) allele key, 3-5) four different methods (DEseq, edgR, McManus, limma) of regulatory classification, 7) allele-specific read counts (e.g. 181203_Barke_limma_results.csv), 8-9) two methods (DESeq and edgeR) of differential expression results, 10) the linear fit results, 11) gene information, 12) limma results, 13) gene count (not allele specific) matrix (e.g. 181203_Barke_limma_results_matrix.csv), 14), norm factors, 15) the sample key, and 16) voom-transformed count matrix.

License: CC BY 4.0 (Creative Commons Attribution)

DOI: 10.5447/ipk/2020/1