How AI helps scientists design the next generation of medicines
View original at technologyreview.comO que extraímos desta fonte
The claims Via News extracted from this document. We point to the source; we don't replace it.
The 'lab of the future' system uses AI to make predictions, robotic systems to execute experiments, and instruments to generate data, analogous to how a self-driving car uses sensors and models to navigate.
60% confidenceHuman oversight remains central even as autonomous AI systems in biologics become more prevalent, ensuring explainable and ethical AI for patient benefit.
60% confidencePredicting whether a computationally generated molecule will be safe in the human body is one of the hardest and least discussed problems in de novo design.
60% confidenceAI-driven models could help identify which two or three disease targets to prioritize and optimize across potency, stability, manufacturability, and safety for multi-specific biologics.
60% confidenceAutomated high-throughput systems will generate AI-ready data at a scale traditional workflows cannot match, with robotic sample handling and integrated data pipelines accelerating early drug development.
60% confidenceVirtual clinical trials using advanced cell systems and micro-scale organ models, paired with AI, can generate enhanced biological safety signals without traditional testing bottlenecks.
60% confidenceAstraZeneca's proprietary, multimodal datasets covering molecular structures, binding measurements, safety profiles, and manufacturing outcomes are the company's key differentiator for training AI models.
60% confidenceScientists will work hand-in-hand with AI model systems, which will design molecules that scientists then test and iterate on.
60% confidenceGenerative AI, combined with other computational tools, could cut drug discovery timelines by as much as 50%.
60% confidenceA shift is underway toward agentic AI systems that simultaneously generate molecule candidates and predict their efficacy and safety, bridging previously separate disease-insight and molecule-design data silos.
60% confidenceAstraZeneca's engineering teams, including data scientists, automation specialists, and AI engineers, are building AI systems that act as transparent 'thinking partners' rather than black boxes, tackling problems like multimodal data fusion, closed-loop optimization, uncertainty quantification, and interpretability.
60% confidenceAstraZeneca's design, make, test, and analyze processes are now computationally enhanced, shortening cycle times while increasing productivity and innovation.
60% confidenceDrugging previously undruggable targets is becoming achievable through AI-driven biologics design.
60% confidenceThe end-state vision for AI in biologic drug discovery is fully AI-generated 'de novo' biologics designed from scratch to a clinical candidate, and this is expected to eventually be achieved.
60% confidenceFuture biologic medicines depend on combining world-class AI and engineering talent with deep scientific expertise.
60% confidenceScientists will remain central to the AI-driven discovery process, providing oversight, judgement, and strategic direction to ensure outputs are explainable and directed toward patient benefit.
60% confidence