We are in the age of the fourth industrial revolution, powered by data and computing and available in public cloud-based computing infrastructure. With this fourth industrial revolution, it is apparent that the world is moving faster. For example, our scientists and doctors have been able to produce a COVID-19 vaccine in a record time of one year or less.  And we have AlphaFold’s open source protein structure and 3D printing, where scientists would be able to weave proteins to form a shape of our artificially created organ, like we do with LEGO® blocks. These discoveries give hope that there will be personalized medicine for each of us based on our individual symptoms and diseases, and there will be a cure for cancer.

In all of the examples cited above, the basic building blocks are the ability to analyze unstructured data at rapid scale, with the power of public cloud (viz. AWS or any other provider) and application of artificial intelligence / machine learning (AI / ML) algorithms. These can produce and detect patterns (e.g., protein structures of molecules present in virus or bacteria or cell structure undergoing evolution inside our body or in affected organs) of various biomarkers, design drug molecules and experiment rapidly in an agile manner to bring speed and scale inside the laboratory.

Scientists need accelerators with high-performance computing resources to record readings from various laboratory instruments, while conducting experiments to convert materials to molecules, store images, and see patterns to detect false positives in rapid scale and speed. AI / ML algorithms can take huge amounts of unstructured data (e.g., images from the laboratory, three dimensional molecular structures, etc.) as input, load them into an artificial neural network, train them to fine-tune and create a proper model with accuracy to detect patterns with lots of computing resources. It requires a lot of investments for organizations to set up such a high-performance computing infrastructure. The good news is that organizations can rent such infrastructure in the cloud at a fraction of the cost and then decommission it when they are done with experiments, without incurring ongoing maintenance cost. For example, data storage and AI / ML tools are available as managed services in-cloud (e.g. AWS rekognition, AWS comprehend medical, S3 or Healthlake) to accelerate analysis to fine-tune material composition to design doses rapidly. Applying to the mouse model in the laboratory creates the refined fine-tuned doses ready for the next stage for clinical trials on human beings.

All these advancements in drug discovery point to the enormous needs to bring the project management discipline with agility inside laboratories. We as project managers can help scientists with AI / ML as a service, by putting together a drug design laboratory in the cloud with the help of technologies from providers (such as Amazon’s AWS health lake) and others for scientists to record data, run tests, detect patterns with AI / ML, and predict experimental results to reduce time and money spent on false positives.  Project managers can guide them to evaluate and negotiate on better terms with cloud-based AI / ML providers, so that drug discovery life cycle is shorter and more affordable in the long run by reducing waste and cycle time.

Here are a couple examples for reference:

‍Moderna on AWS

‍‍Using AI to Accelerate Drug Discovery