Tools for Microbial Genome Assembly and Finishing

De novo genome assembly and genome finishing is becoming increasingly important with the massive amounts of data being generated by next generation sequencing. The enormous amounts of sequencing data created by new sequencing methods still leaves room for Sanger sequencing data in finishing projects for validation and closure. The hybrid data sources can complement each other to generate high quality assemblies. Here we demonstrate how an assembly of a microbial genome can be optimized and closed.

Methods: The publicly available sequence data for Pseudomonas aeruginosa MPAO1 was used to optimize the parameters for genome assembly and finishing. The sequence data is available from the Short Read Archive with accession numbers SRR396636 and SRR396637. The contigs were assembled using CLC bio de novo assembler. To evaluate the assembly, and to design the reagents for genome closure, we used Microbial Genome Finishing Module. This package adds genome finishing and closure functionalities to the CLC Genomics Workbench.

Results: We optimized the word and bubble sizes for this dataset to obtain the initial assembly with the best quality. This step allowed us to reduce the number of contigs from 48 to 23. The contigs were analyzed using the corresponding tool of the Finishing Module. This identified and annotated the problematic regions that needed further attention. Those were the regions with low, high, single-stranded, and unstable coverage. The regions with unaligned ends were also identified and annotated. After manually inspecting and editing mappings in these areas, we aligned the resulting contigs to Pseudomonas aeruginosa PAO1 genome assembly. The alignment of contigs to this close reference identified overlapping contigs, which were merged using the corresponding tool. To close the gaps in the genome, we designed primers for Sanger sequencing using the automated primer design tool. The primers were designed for the ends of all contigs and for the regions with low coverage.

Conclusion: Optimizing de novo assembly parameters allowed us to significantly reduce the number of contigs. The set of genome finishing tools we used produced a higher-quality genome assembly than de novo assembler alone.