The promise of personalized medicine

A new technology developed by scientists at IBM could bring the promise of personalized medicine one step closer to reality.

Using a basic computer language, the researchers created a “smart” DNA stream that contains a patient’s entire medical record, according to a report in the upcoming Oct. 11 print edition of the Journal of Proteome Research, a peer-reviewed journal of the American Chemical Society, the world’s largest scientific society. The report was published online July 22.

With the advent of the genomic revolution, scientists are avidly seeking correlations between human disease and the architecture of individual genes. Parsing this huge amount of data could eventually lead to “personalized medicine,” some researchers say, allowing doctors to prescribe the right drug at the right dose for the right person, based on unique variations in their DNA. But to achieve this potential, scientists need a way to store and efficiently transmit whole sequences of patient DNA with built-in privacy — a hurdle that has yet to be overcome, according to the authors.

Enter IBM’s Genomic Messaging System (GMS). GMS provides a basic computer language that can be inserted into DNA sequences to bridge the gap between patient medical records and genetic information, says lead author of the paper, Barry Robson, Ph.D., a chemist at IBM’s T. J. Watson Research Center in Yorktown Heights, N.Y.

The stream of information transmitted is basically a “smart” DNA sequence containing a patient’s entire medical record in compressed form as well as genetic information. The DNA stream could potentially even house images like MRIs and X-rays. “It is a stream of DNA symbols — GATTACAGATTACA — with GMS language inserted at appropriate points,” Robson says. The inserted language can be used to annotate the DNA, to link to relevant medical data, and to control the privacy of selected sequences with passwords, among others.

Such a universal medical record could help doctors create individualized prescriptions and treatment regimens, precisely tailored for each patient, Robson predicts. “GMS links archives of digital patient records to enable analysis of those records by a variety of bioinformatic and computational biology tools,” says Robson. These tools include data mining to discover unexpected relationships, large-scale epidemiological studies and three-dimensional modeling of patient proteins to study the effect of “SNiPs” — single nucleotide polymorphisms.

Scattered throughout the human genome are millions of one-letter variations in genetic code known as SNiPs. Most are harmless, but some SNiPs provide crucial information, because they can help pinpoint the location of genes that might influence certain diseases.

GMS also provides platforms for respecting the privacy and security of a patient, including a flexible system of passwords that releases only selected parts of the patient’s DNA sequences to different researchers. And since future applications might include medical emergencies, the system has been designed to continue operation even in the event of a disaster by providing a transient backup.

GMS is still in the early stages of development, but in an initial study it successfully modeled SNiPs in proteins from a real patient record. The test, which is one of the first proofs of a fully automated system for personalized medicine, focused on finding and designing a drug that would regulate the rejection of bone marrow in a transplant patient.

Also in earlier research, Robson and his coworkers demonstrated their system’s ability to mine patient data for interesting correlations, such as the connection between a pancreatitis disease and a scorpion bite.