Genome sequence assembly and annotation The blood samples used for genome sequencing were acquired from the Everland Zoo of Korea (Amur tiger, white Bengal tiger, African lion and white African lion) following the Everland Zoo (Korea) ethical guidelines and procedures, and a muscle sample was obtained from a Mongolian snow leopard carcass preserved in the Conservation Genome Resource Bank for Korean Wildlife, Seoul National University. No animals were killed or captured as a result of this study. Libraries for the Amur tiger genome were constructed at BGI, Shenzhen, and the insert sizes of the libraries were 170bp, 500bp, 800bp, 2kb, 5kb, 10kb and 20kb. The libraries were sequenced using HiSeq2000. Other big cat genomes were sequenced at Theragen BiO Institute (TBI), Korea, using HiSeq2000 with read and insert lengths of ~90bp and ~400bp, respectively.The corrected reads were used to complete the genome assembly using SOAPdenovo13. First, the short insert size library (170bp, 500bp and 800bp) data were used to construct a de Bruijn graph. Second, all reads were realigned with the contig sequences. The amount of shared paired-end relationships between pairs of contigs were calculated and weighted with the rate of consistent and conflicting paired ends, before constructing the scaffolds step by step from the short insert size paired ends to the long distant paired ends. Third, the gaps between the constructed scaffolds were closed using the paired-end information to retrieve read pairs where one end mapped to a unique contig while the other was located in the gap region.The Amur tiger genes were predicted using three approaches. First, de novo prediction was performed using the repeat-masked genome using AUGUSTUS (version 2.5.5)32 and GENSCAN (version 1.0)33. Second, homologous proteins in other species were mapped to the genome using tBLASTn (Blast 2.2.23)34 with an E-value cutoff of 1E-5. The aligned sequence and its query protein were then filtered and passed to GeneWise (version 2.2.0)35 to search for accurately spliced alignments. Third, cat EST and full-length cDNA sequences (from UCSC) were aligned to the genome using BLAT36 to generate spliced alignments. For EST results, spliced alignments were linked according to overlap using PASA37. Source evidence generated from the three approaches was integrated with GLEAN38 to produce a consensus gene set. Then, the Amur tiger genome sequence was aligned to two well-assembled and annotated genomes (human and domestic cat) using LASTZ (version 1.02). Finally, mapped results yielding information on homologous proteins were filtered by syntenic blocks of genome sequences. We also predicted the domestic cat (Felis_catus-6.2) gene set, because the gene set of the cat genome is preliminary.
Posted on: Wed, 18 Sep 2013 03:34:19 +0000