We protect human health through microbes. From health across the lifespan to what an ageing society will face — microbial genomics and AI, used together on problems that have stayed unsolved.
Research introduction · From lifting microbes out of a river, to reading them in genome data, and on to human health.
For many years this laboratory has done the work of finding new microbes in nature and giving them names, isolating bacteria never previously described from rivers, soil, algae, salt lakes and even air conditioning systems.
Alongside that came the study of how microbes signal to one another and live in communities, and of the power microbes have to restore contaminated land.
Artificial intelligence now builds on that foundation. The work of lifting a sample from a riverbank becomes the work of finding it first inside public genome data.
Where it leads is human health. A growing body, a working life and old age each need different things. Addressing the health problems particular to each stage of life through microbes — and preparing answers for what the fastest-ageing society in the world will face — is the direction this laboratory is taking.
Find strains that can make what we need without relying on pathogens. The starting point is microbes that people have been consuming safely for centuries.
Gout in men in their twenties to forties, bone and vessels after menopause and in later life, muscle loss and resilience in old age — problems that change with the stage of life, each taken on its own terms.
Korea is ageing faster than any country in the world. What is needed now is not treatment after illness arrives but ways of staying well for longer. Microbes may be the key.
Reverse the order: find candidates in public data first, and use the bench to confirm rather than to hunt. Then write the procedure down so that others can use it too.
Before any experiment begins, we search the public genomes and the patents. Instead of testing a hundred strains blindly, we obtain five or ten with reasons behind them. The difference in cost and time is an order of magnitude.
However well a strain performs, it will not reach industry if its safety is hard to prove. So we start from organisms with a record of being eaten, and build regulatory approval and real-world application into the design from the outset.
Rather than averaging outcomes, we trace why something happens. Collecting results alone leaves you stuck when reports contradict; understanding the mechanism shows you what to change.
Papers on the distribution and origin of the genomic island, and a methods paper on the search procedure itself. Correcting the annotation errors in the public databases is a contribution in its own right.
Candidate production strains that do not depend on pathogens, together with their characterisation. Strains obtained are deposited so that others can use them.
A safe organism means lower verification costs and an opening into food applications. It is the basis for making domestically what is currently imported.
Students are trained to handle data and the bench together — reading a genome, growing the organism, and confirming it by structure, all within one laboratory.
Six topics are under way. They are still at the stage of shaping ideas and gathering evidence, and stay within the laboratory until that work is done.
Grouped by research area. Click any paper to open its topic, methodology and results, each shown with the journal it appeared in.
Microbiology and artificial intelligence, used together for human health.
Reference documents for getting started. Each one can be read directly in the browser, and is also available to download as a PDF file.
Tools running on the laboratory server, ready to use. All four are reached with the same account.