Transcriptome Complexity Among Maize Inbreds

To globally identify expressed PAV genes the transcriptomes of 5 tissues from the 27 NAM founders were deeply sequenced. This RNA-Seq dataset has been analyzed to systematically investigate other types of transcriptomic variation. Consistent with the hypothesis that variation in gene regulation contributes to phenotypic diversity, the genes that exhibit the most variable expression across genotypes are enriched in regulatory functions.  In contrast, genes that exhibit highly variable expression across tissues are not enriched in regulatory functions. Approximately 5 million SNPs and small InDels were discovered in this RNA-Seq dataset. 65% of these variants are located in the transcribed regions of filter gene set (FGS) genes, making this dataset a rich source of genic variation.  Minimum numbers of haplotypes per gene among NAM founders are identified with these genic SNPs.  Consistent with the complementation model of heterosis thousands of FGS genes are affected by Large Effect SNPs (LES), such as premature termination codons, splice site variants etc.  26% of these LESs have allele frequencies less than 0.1. However, an additional 44% of LES have allele frequencies more than 0.8, possibly due to mis-annotation of B73 genes and/or unidentified isoforms. These RNA-Seq data provide evidence for the expression across genotypes and tissues of thousands of working gene set (WGS) genes that are not included in the FGS.  Finally, using stringent cut-offs we identified thousands of novel transcribed regions that are not included in the WGS.