After quality control, the short insert sequences were assembled using the ABySS assembler and the best assembly for each strain was selected for scaffolding and gap filling using the matepair reads.
Sequence data were assembled using NewBler assembler and Consed.
All sequence reads from the P1 sample, including 107,589 Sanger reads and 1,285,077 454 reads, were assembled using Celera Assembler and a large number of the remaining sequence gaps closed through primer walking and PCR closure techniques.
Sequences were assembled using TIGR assembler and additional directed sequencing reactions performed as necessary to complete the sequence to high quality.
The Porphyromonas gingivalis W83 reference (GenBank NC_AE015924), with 2 343 476 bases in one circular contig, had been sequenced by Sanger chemistry, assembled with TIGR Assembler, and finished at TIGR/JCVI (Nelson et al., 2003).
To overcome the under-assembled problem of Mira 3 assembler and enhance the quality of assembling, the unigenes were reassembled using gsAssembler v2.3 with an identity parameter of 95% (Roche NimbleGen, Inc., Madison, WI, USA).
A minimal contig set was assembled using the phrap assembler and all contigs were manually curated.
The 454 virome was assembled with the CAMERA assembler and the Illumina version with IDBA_UD.
WGS3 was assembled with the Celera assembler and hence include a set of degenerate scaffolds (called "armUextra").
The resulting quality trimmed sequences were extracted and assembled with the CAP3 assembler, and viewed with Consed [ 60].
This sequence has been assembled using the JAZZ assembler, and consists of 9,080 scaffolds, containing a total of 1,591,853 reads, and 227.1 Mb.
