Computational Development of a Comprehensive Database of Drug-Drug Interactions
By Amy Tai
In fact, for a time, I really was stuck. For the entire summer between sophomore and junior years, my research project rusted in a corner, because I had not yet discovered the true meaning of “computational” in computational biology. To me, “computational” was still the four major operations: addition, subtraction, multiplication, and division; I was computing numbers, much like a cheap, gas-station calculator. Little did I know, however, that there is an entire field of artificial intelligence focusing on classifiers, statistical learning methods, and intelligent systems . . . An understanding of drug-drug interactions (DDIs) impacts fields ranging from medicine to drug development to public health. In 2004, the average American took a combination of 12 prescription drugs per day [1]. This daily behavior seems trivial, but lack of proper DDI knowledge puts millions of individuals at risk, as a set of 12 drugs could cause more than 1000 lethal interactions. Also in 2004, more than 1% of all deaths were directly caused by DDIs, because these patients were oblivious to the life-threatening reactions that their drug repertoire would cause in the human body [2]. By developing a comprehensive database of DDIs, we can hopefully reduce the number of deaths associated with DDIs . . . There are existing programs that help pharmacists dispense better combinations of drugs, but these are sparse and often inaccurate. One of the goals of this investigation is to improve these programs in accuracy so that pharmacies can be more effective in preventing fatal DDIs…..