But the New York Genome Center, like other scientists studying cancer, has enlisted the help of computers to smooth this difficult path between genetics and treatment.
As reported in an article by Carl Zimmer in the New York Times, using IBM's Watson supercomputer, which in recent years has been "dedicated" to medicine, Dr. Robert Darnell and his colleagues will try to recommend personalized treatments to patients with a deadly type of brain cancer, known as glioblastoma.
Watson currently has the ability to search for medical information in more than 23 million scientific publication abstracts within minutes or seconds.
In this new research by the New York Genome Center, which will initially involve 20 patients, scientists are collecting the genome from biopsies, decoding it and feeding the Watson supercomputer with this data.
Watson will then identify the mutations, consult the medical literature, develop a hypothesis about how they cause cancer, and come up with a list of drugs that could potentially be used to treat the patient.
Some of these drugs may target cancer cells directly, while others target cells that "collaborate" with cancer cells. Some drugs may have been studied for other types of cancer, while others may have never been used in cancer treatment.
A team of experts, including neurologists and pharmacologists, will evaluate Watson's results and make the final decisions about whether to give the patient any of these drugs. Instead of patients starting treatment with just one drug, as is usually the case, Dr. Darnell hopes that participants in this new study will be able to take up to five different drugs at the same time. As he explains, because glioblastoma is a deadly disease, the first signs of whether this method prolongs patients' lives will not be long in coming.
And while several scientists are awaiting the first results before expressing an opinion on this project, they generally agree that supercomputers will be an essential tool in understanding the cancer genome.
Source: tro-ma-ktiko.blogspot.gr
