About half of the genetic variation linked to phenotypic traits is thought to be caused by genetic variation in the non-coding genome, which remains poorly annotated for most species’ genomes. Historically, transcription factor binding sites (TFBS), a key part of the regulatory non-coding genome, had to be arduously identified one transcription factor at a time, or genome-wide through the identification of open chromatin at the cost of interpretability. The novel MNase-defined cistrome-Occupancy Analysis (MOA-seq) provides a scalable approach without compromising on resolution. However, MOA-seq currently relies on the availability of assembled genome sequences, which often remain unavailable. Additionally, sequence quality is paramount for haplotype-specific analysis of TFBS performed in F1-hybrids to avoid trans-effects. This poses a problem both for large and complex genomes, but also for projects with a larger number of genotypes where long-read sequencing often remains unfeasible. We demonstrate the power of MOA-seq for haplotype-specific pan-cistrome analysis in maize, as well as develop novel strategies in barley based on published reference genomes and short-read sequencing. We evaluate these strategies by comparing commonly used SNP-replaced genomes with composite genomes generated based on a practical haplotype graph and genomes generated from long-read sequencing while showcasing the benefits and limitations of each method.
Mardi 5 mai à 11h00
Characterizing the extent and nature of reproductive isolation between wild and domesticated forms: a comparative approach in 14 plant systems
