Molecular Modeling Group of Department of Medicinal Chemistry uses state of the art molecular modeling techniques covering both ligand- and structure-based methods. Variety of applied methodologies is heavily supported by specialized molecular modeling software; several in house applications and scripts were also created to improve and automate repeatable tasks. Panel of professional software suites include: Schrödinger Suite and Accelrys Discovery Studio – molecular modeling tasks, ligand docking, sequence alignment and pharmacophore development, ChemAxon software collection – data management and mining, WEKA and KEEL – machine learning suites. Numerous utilities are used in less challenging tasks, like chemical data conversion and simple analysis (OpenBabel tools) or fingerprints generation (PaDEL Descriptor). In addition data analysis is performed with sophisticated software packages – RapidMiner and Statistica. None of the above listed programs is perfect, and level of complexity of used methodologies often require either manual or scripted ingerentions into predefined workflows or writing tools from scratch. Countless small and big applications were developed in Department of Medicinal Chemistry, using python and third party libraries (mmlibs, openeye chem toolkit), java (ChemAxon plugins) and shell script. Moreover, a highly efficient cluster database engine (ndb) is in use for data aggregation and analysis.
Hardware base of Department of Medicinal Chemistry consist of beowulf cluster of 16 quad core computing units running under control of Linux operating system and efficient queue system SLURM. In addition several powerful workstations are available for research purposes.