| Abstract Detail
Comparative Genomics/Transcriptomics Reynolds, Gillian [1], Mumey, Brendan [2], Strnadova-Neeley, Veronika [3], Lachowiec, Jennifer [1]. Hijacking a metagenomic strategy for rapid comparative subgenomics for polyploids and their progenitors. Polyploidization, or whole genome duplication, is found throughout numerous members of the plant kingdom. Despite this ubiquity, the study of polyploid genomics has been consistently hampered by the sheer complexity they present to sequencing technologies and assembly algorithms. In recent years, newer technologies have largely overcome these issues granting us unprecedented clarity for some of the world’s most important genomes. Now the challenge lies in genomic analysis, with comparative genomics being a core bioinformatic venture for a wide variety of biological disciplines. However, the majority of comparative genomic strategies require a vast amount of computational time and resources, which tend to grow at least linearly with the size of the genomes and the number of genomes to be compared. Such resource requirements can render comparative analysis out of reach, especially given the large size of many polyploid genomes. In this work we present an alternative strategy for comparative subgenomics which enables a top-down, global view of subgenomic sequence content for a broad variety of plant species and enables rapid sequence comparison and visualisation of sequence relationships. To achieve this, we explore subgenomics from a metagenomic standpoint. Where metagenomics has the problem of one sequencing dataset containing many distinct genomes, subgenomics has the problem of one sequencing dataset containing many subgenomes. To that end, we use SourMash, which uses the MinHash strategy for sequence signature generation, to rapidly and efficiently compare the sequence content of subgenomes via clustering for a wide variety of polyploid genomes that are assembled to chromosome-scale quality. Where available, we also explore the subgenome-progenitor relationship. We find a wide variety of subgenomic relationships exhibited by allopolyploid plant genomes in line with those reported by the literature demonstrating the accuracy of this approach. These include subgenomes and progenitors exhibiting a spectrum of evolutionary distances, equal and unequal subgenome degradation and progenitor clustering relationships that reflect their known evolutionary distances given the time of their genome donation. Further, we show that SourMash can produce phylogenetically coherent results and holds promise for being able to rapidly screen for potential progenitor genomes. We explore the limitations of SourMash both biologically and technically, discuss their implications and make recommendations for further implementations of this strategy for polyploid plant genomes.
1 - Montana State University, Plant science and plant pathology, Plant Biosciences Building 305, Bozeman, MT, 59715, USA 2 - Montana State University , Gianforte School of Computing, Norm Asbjornson Hall 253B, Bozeman, MT, 59715, USA 3 - Formerly Montana State University, Gianforte School of Computing, Bozeman, MT, 59715, USA
Keywords: genomics polyploidy bioinformatics.
Presentation Type: Oral Paper Number: CGT2004 Abstract ID:807 Candidate for Awards:Margaret Menzel Award |