Parça bazlı ilaç tasarımında karışık tamsayılı programlama

Developing computational methods for drug design has become one of the most widely studied areas of life sciences in the past two decades. The main objective is to discover new drugs using the information regarding topology and interaction energies of known drugs, that minimizes binding energy and docking energy. Designing new drugs from the fragments of known drugs rather than libraries of larger molecules provides more combinations in products. From these many possible combinations it is possible to nd new drugs that have lower energy values than existing molecular structures. Experimental methods were applied to screen thousands of low-molecular-weight compounds for testing their binding to the target protein in using fragment based drug design effectively. Successful results in fragment based drug design were only obtained with the strong integration of computational techniques, innovative NMR experiments and X-Ray crystallography. Many existing computational algorithms cannot accurately predict the affi nity with which fragments might or might not bind to the protein surface alone. In this work, we developed a computational model which includes both learning and design parts. First, we apply a learning algorithm on known results of docking and binding energies to determine weights for the each fragment reliably. Then, in the design part, the regression model is formed based on the results obtained in the first step, to predict the contribution of each group and individual fragment in docking energy and binding energy and finally designs entirely new molecules based on the fragments available.