Bacterial infection is still a serious threat to humans.  Because of the great diversity of disease-causing bacteria, there is a limitation in the ability of current bacterial identification tests using bacterial culture or antibodies.  DNA sequencing is more suited to identify bacterial species correctly, but the time and cost for sequencing was a problem.  We thus developed a genome analyzing system for the rapid diagnosis of infectious diseases.  Using a nanopore-based, single-molecule DNA sequencer, MinION, and two high-spec laptop computers, we could assemble a system that can in principle identify bacterial species in about one hour from a DNA sample of bacterial infection.

One of the key technologies here was our original database that comprehensively collected genome sequence data.  The database, called GenomeSync, is one of the largest collection of complete genome sequences, containing genomes of 26,223 bacterial species (as of October 2017).  Another essential technology was a software package, called GSTK, for quick and accurate identification of bacterial species for each sequencing read.  This can run sequence similarity searches against the GenomeSync database on a multi-core computer, and can summarize the outputs of similarity searches based on the biological taxonomy.  Furthermore, in order to get the sequencing results as quickly as possible, we installed a software for extracting sequence data from the sequencer outputs while the sequencer is running, which enabled us to analyze sequence data in a real-time manner.

Owing to the successful combination of these core technologies, we could realize the rapid and accurate analysis of bacterial composition.  We hope that our system will be widely utilized for the diagnosis of infectious diseases in the future.