Back to search

BIA-Brukerstyrt innovasjonsarena

An AI platform to facilitate the rapid development of T cell based diagnostic tests: applied to SARS-CoV-2

Alternative title: AI plattform for rask utvikling av en T-celle basert diagnostisk test: anvendes for SARS-CoV-2

Awarded: NOK 7.7 mill.

During the project NOI has been repurposing its proprietary neoantigen prediction platform, the NEC Immune profiler (NIP), to enable it to profile the SARS-CoV-2 virus for immunogenic T cell targets. The repurposed NIP was used to build extensive epitope maps for all the viral proteins across >200 human HLA-alleles. The epitope maps were then analysed using a Monte Carlo statistical modelling approach to identify regions of the virus that are particularly rich in immunogenic T cell epitopes, called hotspots. Hotspots that were homologous to the human proteome and/or were located in non-conserved regions of the virus were removed. Finally, the hotspots were analysed using a digital twin population modelling analysis, which was based on the HLA haplotype of >22,000 people across multiple different geographies, to determine the optimal constellation of hotspots that could be used to develop a diagnostic that was effective in a global population. The output of the analysis was a set of 12 different viral hotspots which could be used to develop a universal diagnostic test that would theoretically cover >95% of the global population. Some of this work was recently published The 12 hotspots were then profiled using NIP to identify individual epitopes that could be presented by the 12 most common class I HLA alleles in the Norwegian population. This analysis resulted in 65 epitopes that were then synthesised and tested on blood from convalescent donors at Professor Ludvig Munthe at Oslo University Hospital. All 65 peptides were shown to stimulate T cell specific recall responses in convalescent sera from a least one donor and most across multiple donors. In addition, 90% of donors responded to at least one of the tested peptides. This data confirms that the viral epitopes predicted to be immunogenic by NIP generated immune responses in donors following infection with SARS-CoV-2. In addition, the data suggests that these peptides could form the basis of a universal T cell diagnostic test.

It has been over a century since we encountered a pandemic like COVID-19. The rate of global transmission and the associated morbidity and mortality have been both shocking and devastating. The pandemic has sparked fears of a chronic worldwide recession or depression. Shut-downs, social distancing and travel restrictions have destroyed many jobs and businesses. A reliable diagnostic test to quantify whether citizens have been infected or have recovered from the virus is critical to overcome this crisis. A proposal that is gaining increasing traction is the use of diagnostic tests to generate immunity passports—i.e., certification that a citizen is immune to SARS-CoV-2. Citizens in possession of an immunity passport could then return to work and travel, thereby reigniting the economy. However, as governments gear up efforts to establish immunity passports based on antibody tests, there is a significant risk that these passports will be denied to immune individuals. Recent studies suggest that only 50% of convalescent COVID-19 patients have a detectable antibody response as their infection has been controlled and eradicated by the cellular arm of their immune system driven by T cells. It is therefore critical to develop diagnostic tests that can quantify cellular immunity towards SARS-CoV-2. In this project NEC OncoImmunity (NOI) will work with our partner Professor Ludvig Munthe at the Oslo University Hospital (OUS) to develop a T-cell diagnostic to complement the antibody tests. NOI will use artificial intelligence (AI) and bioinformatics methods to select the viral peptides that can be used to identify T-cells that provide immunity against COVID-19, which will be subsequently validated using COVID-19 biobanks and T-cell assays developed by Professor Ludvig Munthe’s group. The AI platform developed in the project will guide the development of a novel T cell-based diagnostic for COVID-19 and facilitate the introduction of reliable immunity passports.


BIA-Brukerstyrt innovasjonsarena