New algorithm could improve rare disease diagnosis
Stanford scientists have developed an algorithm, called Phrank, that automates the most labour-intensive part of genetic diagnosis — that is matching a patient’s genetic sequence and symptoms to a disease described in the scientific literature.
Without computer help, this match-up process is said to take 20–40 hours per patient. The expert looks at a list of around 100 of the patient’s suspicious-looking mutations, makes an educated guess about which one might cause disease, checks the scientific literature, then moves on to the next one.
The algorithm developed Gill Bejerano, PhD, associate professor of developmental biology and of computer science at Stanford, and his colleagues is said to cut the time needed by 90%. Findings have been published in the journal Genetics in Medicine.
The algorithm’s name, Phrank — a mashup of “phenotype” and “rank” — hints at how it works. It compares a patient’s symptoms and gene data to a knowledge base of medical literature, generating a ranked list of which rare genetic diseases are most likely to be responsible for the symptoms. The clinician has a logical starting point for making a diagnosis, which can be confirmed with one to four hours of effort per case instead of 20–40 hours. The mathematical workings of Phrank aren’t tied to a specific database, which makes it much more flexible to use.
Phrank is also claimed to dramatically outperform earlier algorithms that have tried to do the same thing, according to the paper. Bejerano’s team validated Phrank on medical and genetic data from 169 patients, an important advance over earlier studies in the field. Prior studies had tested algorithms on made-up patients instead because real-patient data for this research is hard to come by.
“The problem is that this test [using synthetic patients] is just too easy,” Bejerano said. “Real patients don’t look exactly like a textbook description.” On data from real patients, one older algorithm ranked the patient’s true diagnosis 33rd, on average, on the list of potential diagnoses it generated; Phrank, on average, ranked the true diagnosis fourth.
Phrank also holds potential for helping doctors identify new genetic diseases, Bejerano said. For example, if a patient’s symptoms can’t be matched to any known human diseases, the algorithm could check for clues in a broader knowledge base. “You might get the result that mouse experiments cause phenotypes similar to your patient, that you may have found the first human patient that suffers from this disease,” Bejerano said.
Ultimately, “nobody is going to replace a clinician making a diagnosis”, he said. But new technology could help experts use their time more efficiently, helping many more patients get diagnosed, he said.
The lead authors of the paper are graduate students Karthik Jagadeesh, MS, and Johannes Birgmeier, MS. Other Stanford co-authors are Jon Bernstein, MD, PhD, associate professor of pediatrics; undergraduate student Cole Deisseroth; and former graduate students Harendra Guturu, PhD, and Aaron Wenger, PhD. The work was funded by Stanford graduate fellowships, Stanford Bio-X, DARPA, the David and Lucile Packard Foundation and Microsoft. Stanford’s departments of Developmental Biology, of Computer Science and of Pediatrics also supported the work.
Known as Pattern to Knowledge (P2K), the technology uses artificial intelligence to leverage deep...
By analysing the brain activity of tetraplegia patients, a new deep-learning algorithm can help...
Entering the unseen world of atoms, users will be able to follow three possible journeys of the...