In the decades-long search for a vaccine to combat the global AIDS pandemic, one discovery stands out for both its potential benefits and unlikely cast of characters, including a Canadian researcher, a group of Nairobi sex workers and the power of artificial intelligence.
The story begins in mid-1980s Nairobi, where a researcher from the University of Manitoba named Frank Plummer was studying STIs among a group sex workers in the slums of Pumwani. During his research, Dr. Plummer was astounded to find many of the sex workers tested positive for the new virus HIV spreading globally among distinct groups. His team was the first to characterize the continental spread of HIV among women (male to female transmission), which ran counter to popular opinion at the time. HIV was considered a disease of the homosexual population and injecting drug users.
Dr. Plummer, now Senior Adviser at the Public Health Agency of Canada and a 2016 Gairdner Award recipient, discovered that some of the women in the wider study group had a very special quality. Despite being exposed to the virus on a regular basis, many remained HIV negative. “They're basically immune to HIV,” Dr. Plummer told a journalist in 2007. “Their immune systems for whatever reason are able to recognise and kill HIV.”
Dr. Plummer and his team discovered that the HIV-resistant women became more likely to contract the virus after they stopped working as prostitutes or took breaks of two months or more. In essence, they needed to be frequently exposed to the virus in order to maintain their resistivity (natural immunity), a realization that has changed the way researchers think about HIV.
Since the discovery of “natural immunity to HIV”, the mission to identify the mechanism behind the resistivity of these patients has continued, but is limited by the functionality of traditional bioinformatics. This is where artificial intelligence enters the story.
Sightline has been working with Dr. Plummer and his team at the University of Manitoba to develop machine learning algorithms to analyze vast datasets from the Nairobi patients. With the framework in place, the data will soon be fed into MLaaS, which will provide analysis and processing capabilities far beyond anything previously available.
While the short-term objectives of the project are narrow, the potential ramifications are profound. The aim is to establish a computational model for disease diagnostics, which will be of use not only for this specific HIV project, but for disease research around the world. The team will use MLaaS to isolate DNA molecules called single-nucleotide polymorphisms (SNPs) which underlie differences in people’s susceptibility to disease.
The potential benefits are enormous, and could affect our understanding of HIV. The more we can learn about HIV-resistant SNPs, the closer we will be to a breakthrough in HIV research, and potentially a victory in one of the most important public health battles of our time.