Skip to Content

In Silico Immunology

Melvin Cohn uses computers to clarify one of biology’s most confusing fields.
June 1, 2005

For most of us, the jargon of modern immunology is incomprehensible. Even immunologists themselves frequently have trouble understanding their colleagues. So the plainspoken Melvin Cohn is a welcome island in a confusing sea of concepts like self/nonself, costimulatory signaling, and sorting the repertoire. And in fact, Cohn, who at 83 is still a fixture at the Salk Institute for Biological Studies in La Jolla, CA, has clarified his ideas to the point that he has come up with a list of discrete immunologic rules. Those rules allowed him to build a computerized version of the entire human immune system. “If you can’t make it clear, you don’t understand it yourself,” says Cohn.

A microbiologist and biochemist by training, Cohn has spent more than half a century investigating the immune system. He views himself not as a theoretician but as a conceptualist, someone who tries to assemble many small theories into a comprehensible big picture. “The theoreticians largely came from physics and mathematics, and they didn’t understand immunology,” says Cohn, who leads the Conceptual Immunology Group at the Salk Institute. He is well known for elucidating provocative and fiercely debated concepts about the evolution of the immune system and how it operates. “Mel is the smartest, most thorough, most creative living immunologic theoretician,” says Polly Matzinger, an immunologist at the National Institute of Allergy and Infectious Diseases.

One of Cohn’s fundamental immunologic rules, which posits what he calls “protectons,” now serves as the foundation for his computerized “synthetic immune system,” whose roots stretch back to the early 1980s. In 1990, Cohn and longtime collaborator the late Rodney Langman proposed that, just as a brick wall is made of many uniform bricks, the immune system is built of many nearly identical protectons. Each protecton is a corps of about 10 million immune cells and various biochemicals that work together to protect the body from invaders. The bigger an animal, the researchers reasoned, the more protectons it has. The relatively small size of the protecton helped Cohn, Langman, and programmer James Mata deal with the enormous complexity of the immune system when they revamped their in silico model five years ago.

“A mouse has 108 [immune] cells, and a human has 1012,” says Cohn – meaning that a mouse has 10 protectons and a human has 100,000. Yet “a mouse is as protected as a human being,” he says, because it has as many immune cells and biochemicals in each milliliter of its blood. “If you look at a hummingbird of one gram and an elephant of 10 tons, they’re about equally protected from pathogens,” says Cohn.

Available free online, the synthetic immune system still looks somewhat crude, and Cohn acknowledges that “it isn’t user friendly enough,” which he says explains why few people have played with it. Still, “it’s a useful tool that takes what one thinks are the minimal rules of how the immune system works and lets you see what those rules generate,” says Colin Anderson, an immunologist at the University of Alberta in Edmonton, Alberta, whose research focus is on preventing the immune system from attacking organ and cell transplants.

Cohn says the synthetic immune system allows researchers to test various assumptions that undergird their own models of how the real immune system works. He gives the example of a model of how the body defeats an infection that humans normally clear within days. “The computer says it takes a month to mount an effective immune response. Well, the individual would be dead in that time. So which one of these assumptions is causing the delay?”

The University of Alberta’s Anderson says he imagines that as the system improves, it may be able to help answer questions like how to help the immune systems of people with diabetes better tolerate transplants of insulin-producing beta cells. But Cohn says he ultimately thinks the in silico immune system provides a valuable tool, regardless of whether it leads to practical ends. “Understanding has a value in itself, and I’m not going to tell you what use it is,” he says.

For all his plain talk, Cohn has met his fair share of immunologists who don’t have a clue what he’s getting at – or, for that matter, why he has spent more than two decades developing a computerized version of the immune system. But he has learned to live with the doubters. Says Cohn, “I’m so immune to that.”

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.