The challenge for customize drug is to choose the right medicine for the individual patient. Drug testing of patients in large clinical trials is a way of assessing their efficacy and toxicity, but testing hundreds of drugs currently under development is impractical. Therefore, preclinical prediction models are highly desirable because of their capability is to know drug acknowledgement to hundreds of cell lines. In this paper, we presents, the classification technique is used to differentiate the drugs. Initially pre-processing is used on the true data that removes the noise, then the features are extracted which are saved into the database along with the information related to drug response. Then the performance parameters such as precision, recall, accuracy and f-measures are determined. The approximate accuracy observed come out to be 0.9315.
Author(s): Swathi Muppalaneni