De Novo Assembled Transcriptome Analysis and SSR Marker Development of a Mixture of Six Tissues from Lilium Oriental Hybrid ‘Sorbonne’

Lily is universally known for its large, showy, fragrant flower, and Sorbonne is one of the best-selling Oriental hybrid lily cultivars. To accelerate lily breeding for more robust, attractive, and disease-resistant cultivars, it is essential to introduce molecular marker-assisted breeding. However, the enormous genome size, of around 36 Gb, is an obstacle for genomic information mining. We sequenced and assembled a mixture of six tissues of Lilium ‘Sorbonne’ using Illumina HiSeq2000. A set of 49,991 Unigenes with an average length of 673 bp was produced, and 36,093 sequences were annotated against public protein databases. The number of transcripts with GO annotation was 26,212, with 55 subcategories. Virus transcripts were detected, even though in very small amounts. From the 49,991 Unigenes, 1,853 SSRs were identified. The most common repeat motifs were trinucleotide and dinucleotide with GA/CT and GGC/CCG the most common repeats. Fifty-seven EST-SSR markers were validated in 32 lily genotypes with a mean PIC value of 0.55. A 57-primer-based dendrogram was in accordance with the genetic background derived from complex breeding practices and previous studies. Thirty out of the 57 SSR-derived Unigenes were of function-known genes. Subsequently, a comprehensive hybrid assembly was generated, merging the current transcriptome with two more lily transcriptomes available on the public database. The distribution of gene ontology for the three sets of assembly and mismatch/gap uncovering for Unigenes before and after hybrid assembly was analyzed. Our data are an extensive complement to the lily genome database, which will benefit genetic research and candidate gene mining.