Integrating Genome and Transcriptome Data to Predict Growth and Development Traits in Populus deltoides

Gene expression is a major component that determines phenotypic variation observed in natural populations. Uncovering the genes that contribute transcriptionally to trait variation is expected to contribute to our ability to predict phenotypes. Most importantly, models that incorporate transcriptomic information may predict phenotypes more accurately than models based only on DNA sequence variation. To evaluate this hypothesis we previously genotyped and phenotyped an unstructured population of approximately 500 Populus deltoides unrelated individuals, to discover genetic polymorphisms associated with biomass productivity and quality. The genotypic and phenotypic description of this population is being complemented with the characterization of the transcriptome in the differentiating xylem. RNA-sequencing is a powerful approach to capture a snapshot of an organism's transcriptome. Leveraging this tool in combination with existing genomic data, we are now identifying genomic variants that regulated the xylem transcriptional profile. Following this analysis we aim to develop models that will predict the whole plant phenotype using multiple layers of omics data (the genome and the transcriptome). To date we have developed a pipeline that performs quality control, measures and assembles the transcriptome of the association population. Differential expression of transcripts and associations between transcript abundance and phenotypic traits is being conducted. Following these associations, regulatory variants affecting gene expression and phenotypes will be identified.