Key Takeaways:
The 2024 Nobel Prize in Chemistry was awarded for groundbreaking advancements in computational protein structure prediction and antibody design.
Combining GenAI and causal AI enables accurate target identification and antibody design with high affinity, specificity, and minimal off-target effects, while also advancing our understanding of disease mechanisms.
Multi-target identification, with its application for synthetic lethality, could lead to safer, faster, and more targeted cancer therapies.
Each year, the Nobel Prizes celebrate monumental discoveries that push the boundaries of human knowledge. This year’s Nobel Prize in Chemistry is awarded for groundbreaking advancements in computational protein science, awarding David Baker for his pioneering work in computational protein design and Demis Hassabis and John M. Jumper for protein structure prediction [1]. Baker has endeavored to design entirely new antibodies via generative AI (GenAI) [2], potentially transforming the therapeutic antibody market by making antibody design more accessible and cost-effective. Meanwhile, Hassabis and Jumper’s achievements with AlphaFold series of models [3] have solved the long-standing challenge of predicting protein structures from amino acid sequences [4], a crucial step in understanding biological functions and disease mechanisms. Together, these contributions mark significant progress in artificial intelligence in drug discovery (AIDD).
A Synergistic Approach to Drug Discovery: Leveraging GenAI and Causal AI
Despite these amazing progresses, predicting protein structure and antibody design is only half of the story – we still need more effective tools for target identification for drug discovery . On the shoulder of these laureates’ work, Vizuro’s team furthers an dual AI-driven approach by strategically combining diffusion-based GenAI for therapeutic antibody design, and causal AI for revealing disease relatedness and identifying precise targets, pushing the forefront of AIDD. This platform, Apotek Bench , utilizes complementarity-determining region (CDR) grafting and a proprietary CDR optimization framework to enhance the developability of existing antibodies. Thus, Apotek Bench can optimize antibody-antigen interactions and their binding affinity while minimizing immunogenicity, ensuring host compatibility, and potentially improving more effective treatments. On the other hand, Vizuro’s team applies causal AI to unravel hidden causal pathways of molecular entities in multi-omics data, together with the target-centric approach [5], to identify shared targets across diseases. This approach led Vizuro, in collaboration with PharmaEssentia, to present findings at last year’s MPN Congress, revealing novel connections between myeloproliferative neoplasms (MPNs) and neurodegenerative diseases [6]. One of the key results is common inhibitory receptors and inflammatory pathways might underlie both disease types, showcasing causal AI’s capacity to elucidate complex cross-disease relationships.
Collectively, the synergistic approach of GenAI and causal AI offers a robust solution to design antibodies with high affinity and specificity, and potentially with less off-target effects. In addition, it not only advances therapeutic antibody development but also broadens the understanding of disease mechanisms across various conditions.
Multi-Target Identification Is the Next Frontier for Cancer Treatments
One promising application of the aforementioned approach is finding safer, faster, and more effective cures for cancers through multi-targeting. Multi-target identification in cancer treatment represents an advanced strategy that aims to target multiple molecular pathways within cancer cells or their microenvironment [7]. This approach is designed to enhance treatment efficacy and prevent cancer cells from developing resistance by disrupting multiple critical pathways simultaneously. Focusing on multiple targets is particularly beneficial for addressing complex and adaptive cancers, as it reduces the likelihood of treatment escape and improves overall therapeutic outcomes.
A key aspect of this strategy involves leveraging synthetic lethality (SL), where mutations in two separate genes result in cell death, whereas a mutation in only one gene does not [8]. This concept allows for selective targeting of cancer cells while sparing healthy cells. For example, cancer cells with a BRCA1 mutation can be effectively treated by inhibiting the PARP gene, triggering SL and inducing cell death [9]. This approach is instrumental in developing therapies that target specific molecular pairs to enhance the effectiveness of cancer treatments.
Despite its potential, synthetic lethality has yet to see wider adoption due to the challenges in gene pair identification. The current approach still primarily relies on expensive genetic screening, with limited computational methods available. To tackle this challenge, Vizuro’s team utilizes causal AI to identify gene interactions and their impact on cell viability, particularly focusing on pathways such as apoptosis, cell cycle regulation, and DNA repair. By analyzing single-cell RNA sequencing data, Vizuro's team can obtain causal pathways and extract sub-networks formed by SL molecular pairs, offering potential therapeutic targets. This solution aims to advance cancer treatment by integrating these findings into the current AIDD arsenals, allowing for more precise and effective multi-target approaches to combat cancer.
About Apotek Bench
Apotek is Vizuro's comprehensive end-to-end SaaS platform tailored for life sciences. It streamlines drug discovery and research with features like an interactive omics data dashboard, a drug discovery pipeline for immuno-oncology, and multi-dimensional target ranking. Apotek also provides AI-powered bioinformatics support, a digital twin for biosimulation, and services for experimental research. Apotek accelerates scientific insights and innovations in healthcare and biotechnology.
Learn more about Apotek Bench: https://www.vizuro.com/apotek
Reach out to Vizuro team: https://www.vizuro.com/contact
Reference
3 Abramson, J., et al., Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature, 2024. 630(8016): p. 493-500.
4 Dill, K.A. and J.L. MacCallum, The Protein-Folding Problem, 50 Years On. Science, 2012. 338(6110): p. 1042-1046.
5 Parisi, D., et al., Drug repositioning or target repositioning: A structural perspective of drug-target-indication relationship for available repurposed drugs. Comput Struct Biotechnol J, 2020. 18: p. 1043-1055.
7 Doostmohammadi, A., et al., Potentials and future perspectives of multi-target drugs in cancer treatment: the next generation anti-cancer agents. Cell Commun. Signaling, 2024. 22(1): p. 228.
8 Brunen, D. and R. Bernards, Exploiting synthetic lethality to improve cancer therapy. Nat. Rev. Clin. Oncol., 2017. 14(6): p. 331-332.
9 Schäffer, A.A., et al., A systematic analysis of the landscape of synthetic lethality-driven precision oncology. Med, 2024. 5(1): p. 73-89.e9.
Komentáre