Structural bioinformatics: methods development and applications to biomedical problems

In this thesis, I have used a variety of structural bioinformatics methods to investigate different biological and biomedical problems in collaboration with experimental groups, and I have developed a program to help protein structure comparisons. I have combined structure analyses and virtual screenings of several compound libraries to investigate the ligand binding properties of the sigma-1 receptor, which is an attractive therapeutic target for its roles in neural protection and pain. These studies have allowed me to: i) Propose steroid-based compounds as likely physiological ligands sigma-1 receptor; and ii) identify six FDA-approved drugs that were experimentally demonstrated to bind the receptor and improve the growth of cell from Huntington’s disease patients and can be therefore repurposed for use in this disease. I also carried out structure analyses and virtual screening towards COX-2 and COX-1 to compare the predicted binding of compounds with anti-inflammatory potential, obtained from olive oil with eco-sustainable procedures, with classic COX inhibitors. I performed sequence and structure analyses to identify the structural determinants of the biological activity of several Arabidopsis thaliana proteins, in comparison with their human homologues: i) the antiproliferative activity of the two sirtuins; and ii) the substrate specificity of N-acetyltransferase 2. Finally, I have implemented an automated pipeline that compares protein structure alignments produced by multiple programmes and identifies the consensus regions in their output.