Viruses are the most abundant biological entity on the planet. They infect virtually every kind of life form including (sort of) other viruses. Viruses are intensely efficient – some viruses contain as few as 4 genes. Their strategy is typically simple: infect a cell, use its machinery to produce more viruses, and spread to other cells.
Pathogenic human viruses are terrible, but there are many other viruses which are useful for humans. For instance, many modern vaccines use viral vectors to produce antigens of other pathogenic entities. There is also growing interest in using viruses to fight off bacterial infections.
Antibiotic resistant bacteria are on the rise. It has been estimated that 39 million will die of antimicrobial resistant infections over the next 25 years. Bacteriophages (or phages for short) are viruses which infect bacteria. Phage therapy is the use of phages to cure bacterial infections. Phage therapy has existed for around 100 years, but has not become mainstream due to the incredible effectiveness of antibiotics. However, as antimicrobial resistance spreads, the world will need more tools for fighting bacterial infections, and phage therapy is a promising alternative to conventional therapeutics.
One of the biggest proponents of phage therapy in Western medicine is Canadian epidemiologist Steffanie Strathdee. In 2016 Strathdee’s husband, Tom Patterson, nearly died of an antibiotic resistant infection of Acinetobacter baumannii. Incredibly, Strathdee learned about phage therapy and organized a compassionate use application of phage therapy which saved Patterson’s life. The story is documented in her book, The Perfect Predator.
Strathdee recently co-authored a paper titled Optimizing phage therapy with artificial intelligence: a perspective. In it, the authors highlight future uses of AI to improve phage therapy. The most immediate use is matching known phages from phage banks to pathogenic bacteria. This could speed up the discovery of effective phage for a given patient, possibly even when there is no available cell culture of the bacteria.
AI can also be used to suggest mutations to improve the efficacy of a phage. One way to do this is to suggest mutations to the binding region of the phage tail fiber which are likely to increase binding affinity to the bacterial receptors. This can be cast as an inverse folding problem conditioned on the tail-receptor complex structure. However, phage-host specificity is more complex than just receptor binding and so AI may also be used to modify other viral proteins, DNA, or RNA.
The ultimate goal would be to design an entire therapeutic phage genome conditioned on the pathogenic bacterial genome. Such a paradigm would enable targeted therapies for arbitrary bacterial infections which minimally disrupt the beneficial microbiome, using only genomic information derived from a sample such as a blood test. In the meantime, AI can be used as it is in other forms of drug design: narrowing the search space to speed up discovery.
