The Eucalyptus grandisGenome Project: Genome and transcriptome resources for comparative analysis of woody plant biology

Abstract

Background The International Year of Forests 2011 [http://www. un.org/en/events/iyof2011/] will be a milestone for forest tree genomics. The draft genome sequence of Eucalyptus grandis was released in January 2011 in the USA (Phytozome [http://www.phytozome.net]) and in Belgium (BOGAS, [http://bioinformatics.psb.ugent.be/webtools/bogas/]). The genome sequencing was funded by the US Department of Energy (DOE) and performed at the DOE Joint Genome Institute (JGI) in collaboration with members of the Eucalyptus Genome Network (EUCAGEN, [http://www.eucagen.org]) who contributed genetic materials, linkage maps, EST resources and bioinformatics support. The E. grandis genome together with that of Populus trichocarpa[1]and other woody plant genomes recently completed (e.g. Vitis, Cacao, Prunus, Citrus and Malus)will provide excellent opportunities for comparative studies of the unique biology of woody plants. Eucalypts are currently the most widely grown hardwood fibre crop in the world and eucalypt breeding programs will benefit greatly from the new genomic resources. The reference genome sequence of Eucalyptus, a foundation tree genus in Australia comprising more than 70% of the native forest estate, will also offer important benefits for ecological and evolutionary biology studies. We report the sequencing, assembly and annotation of the E. grandis genome. Genome sequencing and assembly Whole-genome (8X) shotgun sequencing was performed for a partially inbred (S1), 17-year-old tree of E. grandis (est. genome size 640 Mbp, n = 11), BRASUZ1 (Suzano, Brazil). A total of 7.7 million Sanger reads (5.4 Gbp) were produced from plasmid, fosmid and BAC libraries. An inbred genotype was selected to circumvent perceived problems with the assembly of a highly heterozygous eucalypt genome. However, microsatellite genotyping showed that BRASUZ1 was much less homozygous than expected, with large parts of the genome remaining heterozygous presumably due to viability selection. This finding was confirmed during the assembly of the S1 genome approximately 25% of the assembly occurred in two haplotypes of 3-4X coverage, while the remainder of the genome assembled into a single haplotype of 6-7X coverage. Linkage maps with over 2400 DArT and microsatellite markers were subsequently used as a framework for the assembly of 11 large chromosome scaffolds. The chromosome scaffolds contained 88% (605 Mbp) of the draft assembly, with the remainder of the assembly sequence (85 Mbp) in 4941 smaller scaffolds. Based on similarity searches with 1.6 million ESTs from BRASUZ1, it was estimated that 96% of expressed gene loci were included in the 11 chromosome assemblies.