While technology alone hasn't evolved as quickly as we wanted, we are now at a crossroads where we have to explore the convergence of technology. How is what happens in robotics is determined by what is happening in the field of 3D printing? What can be achieved by applying the latest advances in quantum computing to nanotechnology? Among all these lines, there is one particularly curious intersection: artificial intelligence and genomics. Each of these areas is making steady progress, but Jamie Metzel believes it is their convergence that will bring us closer to the unexplored territories that we dreamed of reading science fiction. “The tug of war will begin, and it will be a competition between the realities of our biology, with its built-in limitations and the scope of our aspirations,” he says.
Metzel is a senior fellow at the Atlantic Council. Last week he talked about his thoughts on genomics and AI, and where their convergence will take us.
The life we are used to
Metzel explains that genomics as a field has evolved slowly - but quickly gained momentum. In 1953, James Watson and Francis Crick identified the structure of the DNA double helix and realized that the order of base pairs contains a treasure trove of genetic information. It was the book of life, we found it.
In 2003, when the Human Genome Project was completed (after 13 years and $ 2.7 billion spent), we learned that the genome is made up of 3 billion base pairs and the location of specific genes on our chromosomes. The Book of Life didn't just exist - it could be read.
Fifteen years later - in 2018 - we have already mastered the skills to accurately edit the genes of plants, animals and humans. Everything is changing rapidly and pushing us to new boundaries. Forget about reading the book of life - we are learning to write it.
“Read, write and hack - it’s obvious that human beings are just another form of information technology, and just as our information technology has entered an exponential curve of discovery, so will we enter ourselves,” Metzel says. "And that intersects with the AI revolution."
Promotional video:
Life Science Plus Machine Learning
In 2016, DeepMind's AlphaGo program beat the world's best go player. In 2017, AlphaGo Zero appeared: unlike AlphaGo, AlphaGo Zero did not learn from previous go games, but simply learned the rules of the game of go - and beat the AlphaGo program in four days.
Our own biology is certainly much more complex than the game of Go, and that's where we should start. "The system of our own biology that we are trying to understand is massive, but more importantly, comprehensible."
Take the standard set of rules of our biology, the data of the genome - and, in the end, you may even surpass nature itself.
Many countries are already starting to produce such data. The British National Health Service recently announced a plan to sequence the genomes of five million Britons over the next five years. In the United States, the All of Us research program is sequencing a million Americans. China is even more aggressive in sequencing its population and has a goal of sequencing half of all newborns by 2020.
“We will have massive pools of sequenced genomic data,” says Metzel. "The real discovery will come from comparing the sequenced genomes of people with their electronic medical records, and ultimately with their life records."
Getting people to voluntarily share their data is another matter. But this is where the lack of strong privacy protections in China can be a significant advantage.
To compare genotypes and phenotypes at scale - first millions, then hundreds of millions, then billions - according to Metzel, we will need artificial intelligence and big data analytics tools, as well as algorithms that are far superior to those we know. These tools will allow us to move from precision medicine to predictive medicine, so that we know exactly where and when different diseases are ready to occur, and stop their occurrence.
But when we unlock our own genetics, it will be not only and not so much healthcare. Ultimately, it is about who and what we are - people. It's about identity.
Designer kids and their kids
According to Metzel, the most serious application of our genomic knowledge will be in embryo breeding.
Currently, in vitro fertilization (IVF) procedures allow you to extract about 15 eggs, fertilize them, and then carry out genetic testing before implantation; at the moment, you can find out what diseases associated with a mutation of one gene look like, and simple features like hair or eye color. When we get to millions, and then billions, of people with sequenced genomes, we gain insight into how scale genetics work and can make more informed choices.
Imagine visiting a fertility clinic in 2023. You give a piece of skin or a blood sample, and through the use of in vitro gametogenesis (IVG), your skin or blood cells become eggs or sperm, even if you are infertile, which can later be combined into embryos. Tens of hundreds of embryos created from artificial gametes donate several cells, after which these cells are sequenced. The gene sequences tell you about the likelihood of certain traits and diseases. “When the genetic basis is everywhere, we will be able to understand with an ever higher level of accuracy how healthy a child will grow up.”
This, in his opinion, can lead to wild and frightening consequences: if you take 1000 eggs and choose one with the optimal genetic sequences, you can then “marry” your embryo with someone else who has done similar things on a different genetic line. “Your 5-day old embryo and the alien 5-day old embryo will have a baby in the IVG process,” says Metzel. "Then that baby will have a baby with another five day old embryo from a different genetic line, and you can go on and on."
Sounds crazy, doesn't it? But wait, that's not all. As Jason Pontin told Wired this year, “Gene editing technologies like CRISPR-Cas9 will make it relatively easy to repair, add, or remove genes during IVG, eliminating disease or making improvements that ripple through a child's genome. All of this may sound like science fiction, but for those following the research, the combination of gene editing and IVG seems to be more than just a likely - an inevitable development."
Everything crazy is simple
After stepping on the slippery slope of gene editing and embryo mating, we will embark on a dystopian race to create the perfect human. If someone puts so much time and effort into selecting their embryos, Metzel asks, how will he choose a marriage partner for his children? IVG will allow him to accelerate evolution.
“We all need to be part of an inclusive, integrated, global dialogue about the future of our species,” says Metzel. "Health professionals will be important nodes in this." Last but not least, this dialogue may raise the issue of access to technologies such as IVG; Should we take some steps so that the IVG does not go from being a tool for a wealthy minority to a tool for an even richer and more minority, thereby rocking inequality and further polarizing society?
As Pontin notes, 40 years ago IVG also caused fear, confusion and resistance - and today in vitro fertilization is as normal and widespread as the millions of healthy babies created using this technology.
Ilya Khel
