High-Quality, Haplotype-Resolved Assembly and Annotation of an Algal Genome Using Pacific Biosciences Technology

Advances in genome sequencing have driven a rapid expansion of analysis in non-model organisms.  Organisms with high GC content, polyploidy and repetitive elements present especially difficult targets for sequencing.  Understanding ploidy and performing haplotyping to determine allele-specific sequence are also major challenges.  Many current next-generation sequencing technologies are limited in their ability to address these genomes.  We have sequenced a novel 39MB algal genome.  The organism is diploid and has a high GC content, making it especially difficult to sequence.  Initial sequencing attempts using Illumina or 454 produced limited coverage.  We wished to obtain a high-quality reference genome, with haplotype-specific information.  Toward that end, we performed PacBio long read sequencing, combined with PacBio CCS sequencing for error correction.  We then assembled this data using Allora.  The haploid assembly produced was of very high quality, matching the standards of most genomes that have been obtained through Sanger sequencing and laborious finishing efforts. 

We developed an automated haplotyping pipeline which used the PacBio data to accurately convert the haploid assembly into a fully-resolved, diploid assembly, with accurate haplotype phasing.  The resulting haplotype accuracy generally met or exceeded human-derived maps obtained through a combination of Sanger and PacBio sequencing.  We then created a set of custom-trained, haplotype-specific gene models, performed functional annotation of these gene models, and set up a genome browser to view the data.  This work provides several implications for sequencing of difficult genomes in non-model organisms.