Data Mining on Thrombin Dataset

This document describes our work on the task 1 of KDD cup 2001. Different feature selection approaches, different inductive algorithms are employed to analyze the dataset. The results are evaluated and integrated by means of ROC analysis. 1. Background Three tasks in KDD cup 2001 are involved. All of them focus on data from genomics and drug design. The first task is a classification task for a large size of propositionalised dataset, a half-gigabyte-dataset, or thrombin dataset. Both the second task and the third task focused a relational dataset. This documents only discusses the first task. Drugs are typically small organic molecules that achieve their desired activity by binding to a target site on a receptor. The first step in the discovery of a new drug is usually to identify and isolate the receptor to which it should bind, followed by testing many small molecules for their ability to bind to the target site. This leaves researchers with the task of determining what separates the active (binding) compounds from the inactive (non-binding) ones. Such a determination can then be used in the design of new compounds that not only bind, but also have all the other properties required for a drug (solubility, oral absorption, lack of side effects, appropriate duration of action, toxicity, etc.) [1]. Thrombin dataset consists of compounds tested for their ability to bind to a target site on thrombin, a key receptor in blood clotting. 2. Data Understanding There are 139352 features and 1909 instances in thrombin dataset. There are 139351 input features (or predictors), the values of which are 0 or 1. The target feature is ÔActivityÕ, the values of which are ÔAÕ and ÔIÕ. There are 42 ÔAÕs and 1867 ÔIÕs in this dataset.