Skip to Main Content

On-the-Spot Genome Analysis, and Other Jobs for Smartphones

Genome analysis can now be done offline and on the spot, thanks to an algorithm researchers have adapted to perform accurate analysis with less computer memory than current programs.
7m read
Written by:
Photo Credit: Tetiana Lazunova/iStock

One of the great potential benefits of artificial intelligence is the miniaturization of deep analysis. The kind of large-scale data crunching that might have once required a supercomputer, or at least a high-end workstation, can today be done with a lot less processing power on devices small enough to fit in a hand.

A group of researchers recently proved that point by adapting a computer algorithm that allows a smartphone-size device to analyze the human genome all on its own, without having to connect to a heavy-duty workstation or the cloud via an internet connection. The ability for small devices to perform that analysis offline can allow for identification of infectious diseases — such as Ebola, Zika or influenza — on the spot in remote locations or at a hospital bedside, which would accelerate the response and help contain its spread.

The research team from the Garvan Institute of Medical Research and the University of New South Wales in Sydney, Australia, fine-tuned a DNA sequencing program called Minimap2. They were able to use smartphones for the project by paring down the amount of memory needed to align genomic sequences from 16 gigabytes to just two, putting the job within the range of handheld devices, the team reported in ScienceDaily.

“We’re focused on making genomic technologies more accessible to improve human health,” team leader Dr. Martin Smith said. “They’re becoming smaller, but still need to function in remote areas, so we created a method that can analyze genomic data, in real time, on just a mobile device.”

The streamlined algorithm, which researchers said was 99.98 percent accurate in reproducing genetic alignments, is another example of how AI and analytics algorithms are making quick, local work of what used to be big jobs that required the medical computing equivalent of calling in the team with lots of SUVs, sunglasses and heavy blinking equipment.

I’m a Smartphone, Not a Miracle Worker!

Among the popular science fiction tools that have held the most fascination for the tech world is Doctor McCoy’s “Star Trek” Tricorder, a handheld diagnostic tool that could quickly and noninvasively figure out what ailed a person, whether the cause was the flu or some exotic alien bug. A combination of a sensor-laden smartphone and wearable technology is moving the Tricorder closer to reality, but there’s still a long way to go. The best devices so far gather more data than putting a hand on someone’s forehead, but they’re limited to diagnosing about a dozen preset conditions, which is still a far cry what the greenest intern can do.

Nevertheless, as the Garvan/UNSW team showed, a smartphone-size device can accomplish impressive feats that weren’t possible just a few years ago.

The team reduced the amount of computing memory required for genomic analysis by changing the way the algorithm maps readings against a reference library, splitting it into smaller chunks and cutting through the noise by eliminating false and duplicate mappings.

The benefits of on-the-spot diagnoses could be significant. The 2014 to 2016 outbreak of the Ebola virus in West Africa — the largest epidemic of its kind, resulting in nearly 29,000 cases in the three hardest-hit countries and more than 11,000 deaths — at first went undetected in remote regions, which contributed to its spread. Domestically, the ability to detect an infectious disease quickly at a patient’s bedside could help prevent an epidemic, whether it be a strain of flu, SARS, or, gulp, the plague.

“The potential of lightweight, portable genomic analysis is vast,” Smith said. “We hope that this technology will one day be applied in the context of point-of-care microbial infections in remote regions or in doctors’ hands at the hospital bedside.”

Where the AI Meets the Road

Aside from disease detection, smartphones are proving to be simple, low-cost solutions to a variety of problems in the public sphere, even if they work with an internet connection rather than offline.

The nation’s crumbling infrastructure is one area they can be used to help. Aside from empty promises and gridlock in Washington, part of the problem is the time and expense of getting an accurate handle on the state of roads, bridges and other features. Researchers at the University of Missouri have a plan to put smartphone technologies to work on it.

They’ve developed a crowdsourcing approach that would let people use the sensors on their smartphones to easily transmit information to a database while cruising (or bumping, as the case may be) along a road. The built-in gyroscopes, accelerometers, cameras, along with external sensors such as infrared sensors, can combine to produce an accurate, real-time assessment of a road’s condition, the researchers say.

“Assessing roads, bridges and airfields with affordable sensors, such as those found in smartphones, really works,” said research team member Bill Buttlar. “In a recent project sponsored by the Missouri Department of Transportation, we also showed that it can accurately assess the condition of airport runways and taxiways.”

Smartphones also are being applied in other scientific and medical fields. At Florida Atlantic University, for example, a new imaging algorithm can use a smartphone photo to analyze interactions between matter and electromagnetic radiation that typically are evaluated via spectroscopy, which requires a large piece of equipment. And in examining 10,000 images so far, the research team said, the smartphone algorithm has outperformed existing spectroscopy algorithms.

AI algorithms, along with more powerful handheld processing, is miniaturizing the work of taking the pulse, so to speak, of subjects as varied as human patients and the infrastructure. Tricorders on the Star Trek level might not be right around the corner, but Bones probably would at least recognize these new tools, and what they’re trying to accomplish.

Related Content