Biological computing: I know I like this

IMAGE: OpenAI’s DALLА·E, via ChatGPT

As a graduate in Biology and a doctor in Information Systems, this interview on Ars Technica con Jacob Uszkoreitone of the authors of the mythical paper titled come onAttention is all you needcome on what gave rise to generative artificial intelligence and now co-founder of InceptiveI necessarily had to like it: in it, Jakob makes an introduction to the concept of Computación biol a3icaand provides a few very interesting points to draw from to get an idea of ​​the progress in that fascinating field.

Biological computing is a frontier where biology and artificial intelligence come together, a concept that seeks to program living systems through molecules designed as sequences of ribonucleic acid (RNA) that execute specific functions within organisms by being converted into proteins through ribosomes, in a very similar way to how the software directs the operation of a computer. We are talking about an approach that promises to revolutionize medicine and biotechnology, transforming them into much more precise and programmable disciplines.

Jakob is one of the pioneers in this field: he founded Inceptive after leaving Google to apply the deep learning and high-speed biochemical experimentation to programmable drug design. The objective is to move towards what he calls آ«software “biological”, a concept that can completely redefine our relationship with biology and health. That Jakob, like the rest of the authors of that famous paperhad to leave Google to develop biological computing in a new company when Google already had a company like Verily It's one of those things that demonstrates the company's pathetic management practices.

He software biological follows a conceptual model similar to that of the development of software traditional, but with a completely new approach: the desired behavior is defined as the biological function that you want to achieve, such as, for example, activating the production of a specific protein or inhibiting a particular gene. Instead of a conventional compiler, an artificial intelligence-based compiler is used: this model, which is trained with huge volumes of biological data, translates the desired specification into a functional molecule, such as an RNA sequence. Finally, once the molecule is designed, it is introduced into the organism or cell where the cellular machinery is activated to execute the programmed function, just as if it were a computer program running on a biological operating system. .

An example cited by Jakob himself in the interview is the RNA vaccines against COVID-19 by companies such as Pfizer-BioNTech (I got my sixth dose yesterday) or Moderna (my first three). These vaccines contain a molecular program that causes cells to produce a modified version of the virus spike protein, which then triggers an immune response that teaches the body to fight the real virus. But this case is just a relatively simple example. In reality, the potential of software Biological goes much further: molecules capable of regulating genes ultra-precisely, acting on specific types of cells or executing complex behaviors for a certain time.

The inspiration, in reality, comes from the RNA viruswhich with a simple genetic sequence are perfectly capable of replicating, distributing and adapting in organisms and the environment. If we were able to design molecules that mimic a fraction of that complexity for therapeutic rather than pathogenic purposes, we could certainly hope to completely transform medicine as we know it.

The concept of biological computing raises very interesting opportunities in areas such as data management and analysis, because biochemical experiments generate massive amounts of data that require very robust computing infrastructure for management. Bioinformatics and models of machine learning They are key to analyzing this data and training artificial intelligence compilers capable of designing effective molecules. Additionally, computational tools make it possible to simulate molecular interactions and predict how cells will respond to biological programs, thus reducing the need for extensive testing. Being able to understand biological networks as the integrated systems that they are allows us to identify therapeutic targets, and anticipate possible side effects or unwanted interactions.

The potential is enormous. But logically, biological computing faces challenges such as the fact that living systems are nonlinear and emergent, which in many cases makes it difficult to predict the results of a biological program with absolute precision. In this area, it is fundamental and essential to be able to guarantee that the designed molecules are safe and effective. Ensuring the stability and efficient delivery of engineered molecules in the organism is a crucial challenge, and to complicate matters further, the design of complex biological programs requires very high quality data, which are not always available or easy to obtain. Furthermore, the ability to program living systems clearly raises significant ethical dilemmas about the possible scope that this technology could have.

With all this in consideration, my impression is that biological computing can bring us revolutionary advances, from the total personalization of medicine with treatments designed specifically for the genetic and biological characteristics of each patient, to synthetic biology, as organisms designed to, for example, produce biofuels, sustainable materials or repair damaged ecosystems. They can also help us, as in fact they have already managed, in rapid responses to pandemics, through vaccines and therapies that can be designed and produced in weeks against new pathogens (while the most ignorant, who otherwise If it were because of those vaccines, they would all be dead by now, they would absurdly shout to the heavens). Being able to develop biological computing in an advanced way would finally allow us to explore and understand vital processes with a precision that we have never been able to achieve.

Deep down, we are machines, and as such, we are programmable. Biological computing is not only capable of redefining how we understand and treat diseases, but also of opening a whole new era in which biology will be able to be programmed as if it were software. For those of us who combine knowledge in biology and information systems, we are talking about a unique opportunity to lead a technological and scientific revolution in which the key is to collaborate between disciplines, keep up to date with advances in artificial intelligence and biotechnology, and work with the necessary ethical and scientific rigor that this technology demands and requires. Because ultimately, we will be programming not only molecules, but the very future of life and health, with everything that this may entail.


This article is also available in English on my Medium page,آ آ«Biological computing: now that’s a combination I can get behindcome on

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button