Periodic Reporting for period 1 - APPEAL (Antivirus Pandemic Preparedness EuropeAn pLatform)
Berichtszeitraum: 2024-01-01 bis 2025-06-30
Developing new treatments for viruses—especially those identified as high-priority by the WHO—remains critical. Viruses constantly evolve, particularly in vulnerable patients like those with chronic infections, often developing resistance to antiviral drugs. One promising approach is to target proteins in human cells that viruses depend on to survive, rather than proteins of the viruses themselves. This strategy can block the virus without harming the host. A well-known example is an HIV therapy, which targets both viral proteins (reverse-transcriptase, integrase, viral protease) and human proteins (surface receptors for its entry) to stop the virus from spreading. Ideally, such treatments would work against a wide range of viruses, similar to how some antibiotics fight multiple types of bacteria.
How We Find New Targets
Recent advances in functional genomics, such as CRISPR/Cas9 and RNA interference, allow scientists to systematically identify these "host dependency factors" (HDF). By knocking out one gene at a time, researchers can see if the cell becomes more resistant to infection—indicating a potential drug target. This experiment can be done on a large scale covering the complete human genome. However, such genome-wide knockout screens are challenging due to inconsistent results. Furthermore, drugs have to be selected which are efficiently inhibiting these HDF. These are ideally drugs, which are already in clinical use for another disease or indication, as these so called repurposed drugs been already approved (for the other indication) and their safety confirmed, speeding up their use against viruses.
Our Project: Building a Pipeline for Antiviral Drugs
Our goal is twofold:
1. To create a computational and experimental pipeline to identify and validate antiviral drugs.
2. To find at least one broad-spectrum antiviral drug effective against emerging and re-emerging viruses.
We’re using several strategies:
• Machine Learning: We analyse data from gene knockout/knockdown experiments, protein interactions, and gene expression profiles to pinpoint potential HDF.
• High-Density Cell Arrays (HD-CA): These allow us to conduct up to 20,000 experiments on a single slide, providing detailed insights through cell imaging.
• Human Primary Cell Cultures: Unlike cancer cells, these better reflect how viruses interact with healthy human cells.
• Activating Host Defences: We’re identifying host factors that naturally fight viruses and developing drugs to boost their activity using small activating RNA technology.
Our pipeline has two tracks:
• Expedited Arm: We select and test repurposed drugs for broad-spectrum efficacy, moving quickly to clinical trials.
• Comprehensive Arm: This combines all strategies to develop a sustainable system for rapid response to new viral outbreaks.
How We Find New Targets
Recent advances in functional genomics, such as CRISPR/Cas9 and RNA interference, allow scientists to systematically identify these "host dependency factors" (HDF). By knocking out one gene at a time, researchers can see if the cell becomes more resistant to infection—indicating a potential drug target. This experiment can be done on a large scale covering the complete human genome. However, such genome-wide knockout screens are challenging due to inconsistent results. Furthermore, drugs have to be selected which are efficiently inhibiting these HDF. These are ideally drugs, which are already in clinical use for another disease or indication, as these so called repurposed drugs been already approved (for the other indication) and their safety confirmed, speeding up their use against viruses.
Our Project: Building a Pipeline for Antiviral Drugs
Our goal is twofold:
1. To create a computational and experimental pipeline to identify and validate antiviral drugs.
2. To find at least one broad-spectrum antiviral drug effective against emerging and re-emerging viruses.
We’re using several strategies:
• Machine Learning: We analyse data from gene knockout/knockdown experiments, protein interactions, and gene expression profiles to pinpoint potential HDF.
• High-Density Cell Arrays (HD-CA): These allow us to conduct up to 20,000 experiments on a single slide, providing detailed insights through cell imaging.
• Human Primary Cell Cultures: Unlike cancer cells, these better reflect how viruses interact with healthy human cells.
• Activating Host Defences: We’re identifying host factors that naturally fight viruses and developing drugs to boost their activity using small activating RNA technology.
Our pipeline has two tracks:
• Expedited Arm: We select and test repurposed drugs for broad-spectrum efficacy, moving quickly to clinical trials.
• Comprehensive Arm: This combines all strategies to develop a sustainable system for rapid response to new viral outbreaks.
Our progress so far
• We’ve compiled and analysed data from over 90 gene knockout/down screens of cells infected with high-priority viruses, using machine learning to identify potential HDF.
• We’ve optimized drug selection to maximize on-target effects (hitting HDF) while minimizing off-target effects.
• For the first time, we’ve conducted genome-wide knockout screens for highly hazardous viruses like Lassa and Nipah, under the strictest safety conditions.
• We’ve successfully implemented CRISPR/Cas9 technology in high-density cell arrays and adapted it for human primary cells we use, reducing reliance on cancer cell models.
• We’ve generated promising lists of HDFs and drugs, which are now being tested for efficacy against a broad range of viruses of concern.
• We’ve compiled and analysed data from over 90 gene knockout/down screens of cells infected with high-priority viruses, using machine learning to identify potential HDF.
• We’ve optimized drug selection to maximize on-target effects (hitting HDF) while minimizing off-target effects.
• For the first time, we’ve conducted genome-wide knockout screens for highly hazardous viruses like Lassa and Nipah, under the strictest safety conditions.
• We’ve successfully implemented CRISPR/Cas9 technology in high-density cell arrays and adapted it for human primary cells we use, reducing reliance on cancer cell models.
• We’ve generated promising lists of HDFs and drugs, which are now being tested for efficacy against a broad range of viruses of concern.
For the first time, we’ve conducted genome-wide knockout screens for highly hazardous viruses like Lassa and Nipah, under the strictest safety conditions. These screens led to new lists of HDF enabling compound selection and testing, and may lead to new treatments against these viruses. Furthermore, we developed the new software "Slimformer". Omics analyses often yield overwhelming gene lists, making it difficult to extract meaningful biological insights. Traditional methods rely on Gene Ontology hierarchies, ignoring semantic context in gene set descriptions. Slimformer leverages foundation models from Natural Language Processing (NLP) based on gene set names, descriptions, and gene lists. It categorizes gene sets using manually curated molecular processes for benchmarking, relevant for virology achieving excellent prediction performance. "Slimformer" automates and refines gene set categorization, enabling researchers to quickly identify relevant cellular mechanisms from complex omics data. By revealing previously overlooked pathways (e.g. cell cycle dysregulation in RSV), it could fast-track therapeutic target identification and improve understanding of disease mechanisms. It is available via a user-friendly web interface making advanced omics analysis accessible to a broader scientific community. "Slimformer" transforms raw omics data into actionable biological insights, with the potential to advance research in virology, drug discovery, and beyond.