<p>Malaria is a widespread disease caused by parasites of the genus <i>Plasmodium</i>. Each year 500 million clinical cases are reported resulting in over one million casualties. The most lethal species, <i>P. falciparum</i>, accounts for ~90% of the fatal cases and has developed resistance to chloroquine. The resistant strains are a major problem and calls for novel drugs.</p><p>In this thesis, the process of computational inhibitor design is illustrated through the development of <i>P. falciparum</i> aspartic protease inhibitors. These proteases, called plasmepsins, are part of the hemoglobin degradation chain. The hemoglobin is degraded during the intraerythrocytic cycle and serves as the major food source. By inhibiting plasmepsins the parasites can be killed by starvation.</p><p>Novel inhibitors with very high affinity were found by using a combination of computational and synthetic chemistry. These inhibitors were selective and did not display any activity on human cathepsin D. The linear interaction energy (LIE) method was utilized in combination with molecular dynamics (MD) simulations to estimate free energies of binding. The MD simulations were also used to characterize the enzyme–inhibitor interactions and explain the binding on a molecular level.</p><p>The influence of the partial charge model on binding free energy calculations with the LIE method was assessed. Two semiempirical and six <i>ab initio</i> quantum chemical charge derivation schemes were evaluated. It was found that the fast semiempirical charge models are equally useful in free energy calculations with the LIE method as the rigorous <i>ab initio</i> charge models.</p>

Malaria is a widespread disease caused by parasites of the genus Plasmodium. Each year 500 million clinical cases are reported resulting in over one million casualties. The most lethal species, P. falciparum, accounts for ~90% of the fatal cases and has developed resistance to chloroquine. The resistant strains are a major problem and calls for novel drugs. In this thesis, the process of computational inhibitor design is illustrated through the development of P. falciparum aspartic protease inhibitors. These proteases, called plasmepsins, are part of the hemoglobin degradation chain. The hemoglobin is degraded during the intraerythrocytic cycle and serves as the major food source. By inhibiting plasmepsins the parasites can be killed by starvation. Novel inhibitors with very high affinity were found by using a combination of computational and synthetic chemistry. These inhibitors were selective and did not display any activity on human cathepsin D. The linear interaction energy (LIE) method was utilized in combination with molecular dynamics (MD) simulations to estimate free energies of binding. The MD simulations were also used to characterize the enzyme–inhibitor interactions and explain the binding on a molecular level. The influence of the partial charge model on binding free energy calculations with the LIE method was assessed. Two semiempirical and six ab initio quantum chemical charge derivation schemes were evaluated. It was found that the fast semiempirical charge models are equally useful in free energy calculations with the LIE method as the rigorous ab initio charge models.