The accurate description of proteins and the reactions catalysed by them is of uttermost importance for the description of and interference into biochemical processes. Therefore, in this project, new developments improving the accuracy of a fragment-based quantum chemical approach will be performed. In the adjustable density matrix assembler (ADMA) method, fragment electron densities are combined to approximate the electron density matrix of the complete macromolecule. The fragment densities are thereby taken from calculations on smaller molecules including parts of the protein. A general problem in these fragment-based as well as in mixed quantum mechanical / molecular mechanical (QM/MM) approaches is the description of the border regions of the fragments. The first part of the project will therefore be designated to the design of better solutions to this problem. In the second part, the fact that in most cases only a small region of the biomolecular system is involved in the chemical reaction will be used. For the more accurate description of the reaction, larger basis sets and higher levels of theory can be applied to only this central part hopefully reaching chemical accuracy. The rest of the system is still treated by the lower level of theory leading to only a moderate demand of additional computer resources. Additionally, we hope that our research can also give new impulses to the QM/MM field due to the similarities of the approaches used therein.
