DNA Is Better at Math Than You

I was tempted to write about the work being done at Wharton that suggests that AI may already be better at being entrepreneurial than most of us, and of course I’m always interested to see how nanoparticles are starting to change health care (e.g., breast cancer or cancer more generally), but when I saw what researchers at China’s Shanghai Jiao Tong University have done with DNA-based computers, well, I couldn’t pass that up. 

Credit: Bing

If PCs helped change the image of computers from the big mainframes, and mobile phones further redefined what a computer is, then DNA computers may cause us to one day – in the lifetime of some of you -- look back at our chip-based devices as primitive as we now view ENIAC.

It’s been almost 30 years since Leonard Adleman first suggested the idea of DNA computing, and there’s been a lot of excitement in the field since, but, really, not the kind of progress that would make a general purpose DNA computer seem feasible.  That may have changed.

At the risk of introducing way too many acronyms, the Chinese researchers claim they have developed a general purpose DNA integrated circuit (DIC), using “multilayer DNA-based programmable gate arrays (DPGAs).”  The DPGAs are the building blocks of the DIC, and can be mixed and matched to create the desired circuits.  They claim that each DPGA “can be programmed with wiring instructions to implement over 100 billion distinct circuits.”

They keep track of what is going on using fluorescence markers, which probably makes watching a computation fun to watch. 

One experiment, involving 3 DPGAs and 500 DNA strands, made a circuit that could solve quadratic equations, and another could do square roots.  Oh, and, by the way, another DPGA circuit could identify RNA molecules that are related to renal cancer.  They believe their DPGAs offers the potential for “intelligent diagnostics of different kinds of diseases.”

DNA tracking DNA.

"Programmability and scalability constitute two critical factors in achieving general-purpose computing," the researchers write. "Programmability enables specification of the device to perform various algorithms whereas scalability allows the handling of a growing amount of work by the addition of resources to the system."  The authors believe they’ve made significant progress on both fronts. 

Moreover, they say: “The ability to integrate large-scale DPGA networks without apparent signal attenuation marks a key step towards general-purpose DNA computing.”  

I don’t pretend to understand the chemistry, engineering, or computing logic involved in all that, and I’m not saying you’ll soon be carrying around a bunch of DPGAs instead of your phone.  But I’m pretty sure at some point in the foreseeable future we’ll not be carrying around phones as our devices, and I suspect there’s a pretty good chance that DNA is going to be crucial to our computing future.

For one thing, the storage in DNA is unrivaled. As MIT professor Mark Bathe, Ph.D. told NPR: “All the data in the world could fit in your coffee cup that you’re drinking in the morning if it were stored in DNA.” It’s hard to get our heads around how much more efficient – and resilient -- nature is with DNA data storage than anything we’ve come up with.

For another, as long as we’re DNA-based creatures, it’s going to be relevant to us, whereas I already have computer storage disks I don’t have ports for and computers that are so out-of-date as to be useless.  DNA isn’t going to go out of date.

For a third reason, our current approach to computing rely heavily on a wide range of materials, especially the so-called rare earth elements.  It’s not so much that they’re rare as it is that they are incredibly hard to mine and process, and create a significant amount of pollution along the way.  A computing future based on our silicon chip approach is not sustainable and probably won’t survive the 21^st century. DNA is literally everywhere.

Fourth, biology – specifically, brain cells -- brain cells – may be the best path forward to AI, as suggested by a new field called Organoid Intelligence (OI).  “Computing and artificial intelligence have been driving the technology revolution, but they are reaching a ceiling,” said Thomas Hartung, the leader of the initiative that established OI. “Biocomputing is an enormous effort of compacting computational power and increasing its efficiency to push past our current technological limits.”

Professor Hartung pointed out that only last year a supercomputer exceeded the computational capacity of a single human brain — “but using a million times more energy.”

Organoids at work. Credit: Smirnova, et. alia.

Fifth and most specific to health care, we are biological, DNA-based beings, and there’s just something fitting about using biological computing as one of, and perhaps the primary, approaches to how we track and manage our health.  As I wrote several years ago, what could be better than being your own medical record?   

Sixth and finally, we’ve had a great run with our current approach to computing, but it is overdue for the next big thing.  That next big thing may be DNA/biological computing, or it may be quantum computing, or it may be a combination of both, but I would be willing to bet that 22^nd computing doesn’t look much like 2023 computing.  We need to be looking ahead.

So, yeah, I’m excited by DNA/biological computing, and I think you should be too.