Pattern Recognition in Drug Discovery

This is the future of drug development. We no longer have to wait for someone to stumble across the healthful effects of bread mold. Many technologies have come together — DNA sequencing, pattern recognition, and databases of medical records — that could create a surge in medical progress.

For example, AbCellera Biologics’ (ABCL) “full-stack, artificial intelligence-powered antibody discovery platform searches and analyzes the database of natural immune systems to find antibodies that could be developed as drugs.” Harvard University owns 4.1 million shares. ABCL even turns a profit.

The main barrier to this kind of study is access to data. Most of us cannot even assemble our own patient data in one place. Providers have a strong incentive to keep the data: They would rather tell you the results over the phone than add the data to your comprehensive patient record. When data is centralized, individual providers and networks lose control. Providers also fear a trail of responsibility, and patients are taught to fear the possibility of someone learning their medical conditions.

The benefit of centralized medical data, though, is vast. Someday, you will hire a DoctorBot that has looked at 500 million comprehensive patient records from birth to death with personal habits, DNA, medical imaging, and FitBit recordings. DoctorBot will see patterns that no human being could recognize, and suggest, for instance, “You may develop Parkinson’s. Fortunately, we have a pre-treatment for that. Would you like to investigate further?”