My exploration in medicine began with the death of my grandpa due to cancer. He used to drop me off from school in my sixth grade, and within hours of doing that on one fateful day, he left us forever. Right after that, I bought an oncology book. Science was something that I had liked a lot since my childhood, and it seemed to be a tool to seek answers.
Several years down the line, I was the doctor performing physical exams, but I was the same person searching for answers. The hospital where I worked had the privilege of treating many children with rare diseases, almost all of which had no treatment. I realized each day there how much medicine could tell about my patients' illness: the cellular malfunction that I could not understand.
All of us have gone through a similar feeling of anguish and despair. Loved ones are wrongly diagnosed, and prescribed medication that fails to address their needs. Even the language required to discuss such illnesses as Alzheimer’s and Parkinson’s has remained stubbornly unevolved in the face of decades of research that’s yielded nothing so far.
However, I’m convinced all of this can be transformed—far sooner than in 50 years—from a grim reality into a bright hope through concerted efforts of the scientific community toward realizing the enormous potential of artificial intelligence in human health.
One example out of many is the emergence of frontier AI models able to synthesize entirely novel proteins to attack cancerous cells and prevent infection with a range of pathogens. Such AI models work because they have processed large amounts of data, and gained an understanding of the way proteins behave in the human body. The capability to model an entire cell, organ, tissue, or even human biology in its totality should be there too.
Given these prospects, which are rapidly coming to fruition, it is high time for the technology and scientific communities, as well as philanthropic organizations, to lay the groundwork for the next stage of scientific advancement and the discovery of cures for various maladies. No single entity can achieve this alone. This is the reason why we need to unite the entire community and build a foundation for open data in order to drive AI-based biological science forward. That is why my organization, Biohub, has launched the Virtual Biology Initiative.
Robust cell models might revolutionize the discovery process. Over hundreds of years, science has progressed via reductionism. We eliminate all confounders, simplify, and reduce our question to something testable in a lab environment and comprehensible within the confines of a grant cycle. The knowledge we have accumulated is not reflective of human biology.
Unlike the current scientific framework, however, there are no such limitations imposed on AI models. That is why AI promises to provide the scientific community with the tools necessary to tackle the most complicated questions in healthcare. In principle, if AI models are capable of simulating and understanding the workings of the immune system, then it must be possible to develop therapies capable of preventing cancers, neurodegeneration, and metabolic disorders. From what we currently know, the number of potential cures will be dictated exclusively by the size of the model.
But this brings us to the greatest challenge the industry must confront. To train AI models, one must first have data. Protein models are usually trained based on protein datasets. Genomic models are usually built using genomic datasets. Therefore, one would also require a cell dataset and the corresponding model, which, as of now, does not exist.
To get things rolling, scientists will need to work together like never before.
In the last decade, universities and research institutes across the globe have collaborated to expedite scientific understanding of cellular biology, including efforts to support large-scale data generation initiatives such as benchmarking cell maps of humans and other species. We’ve built libraries of cellular imaging data, one of the largest single-cell data sets in the world. In collaboration with the public and private sectors, last year, we launched the Billion Cells Project, a network creating one of the largest ever open-source biological datasets.
This will be the foundation for the Virtual Biology Initiative. For example, to catalyze global coordination, the program launches with a $100 million commitment to fund data generation across the entire scientific community. Other institutions working together with Biohub include Allen Institute, Arc Institute, Broad Institute, Wellcome Sanger Institute, as well as consortia like the Human Cell Atlas and the Human Protein Atlas to coordinate a more comprehensive effort. We’ve also partnered with NVIDIA, along with Renaissance Philanthropy, to catalyze funding for these efforts.
We will continue to develop frontier technologies that enable the measurement of cells at Biohub. For example, part of the $400 million committed towards imaging technology development involves microscopy technology for observing billions of living cells, as well as cryo-electron tomography to capture atomically resolved images within the cell. Cell and tissue engineering capabilities will also be a focal point, enabling scientists to conduct new experiments and measurements on previously inaccessible biology.
If you are capable of funding or performing biology research, I encourage you to take part in this research mission. I am certain that artificial intelligence models will unravel the biological mysteries that remained unsolved in the previous hundred years. By cooperating, we will get the answers much sooner than expected.
Technology of an open-source character makes possible a completely new research process that allows bringing different kinds of specialists together to solve problems no organization can tackle by itself. The concept of personalized medicine has been discussed for many years. This research and the artificial intelligence models it will produce will turn this dream into reality.
Be sure, there are millions of patients out there who have been waiting for us to succeed.
Post a Comment