In silico design of biologics


SNIC 2016/1-506


SNAC Medium

Principal Investigator:

Samuel Flores


Uppsala universitet

Start Date:


End Date:


Primary Classification:

30110: Immunologi inom det medicinska området




We wish to apply for a medium size CPU time allocation of 40000h/month to continue the work of projects SNIC 001/12-240, SNIC 2013/1-302, SNIC 2015/1-443, and SNIC 2016/7-75. Due to recent efficiency improvements we can accept a somewhat smaller allocation, e.g. in the range of 20000-30000 hours if needed. We would like for this allocation to begin immediately rather than waiting for SNIC 2016/7-75 to expire, as the latter is a very small allocation. Our method called ZEMu (Zone Equilibration of Mutants) offers a strategically important means of designing biologics (monoclonal antibodies or alternative scaffolds) in silico. It accounts for backbone flexibility by employing one or more small flexibility zones in internal coordinates. Then it computes dynamics taking into account physical interactions only within the physics zone -- a small spatial neighborhood about each flexible zone. We published papers on this method in PROTEINS, Nature Scientific Reports, PLoS Computational Biology, Human Mutation, and Molecular Biology and Evolution, and acknowledged SNIC. We are evaluating the compatibility of ZEMu with potentials other than FoldX, to increase the efficacy, convenience, and impact of the method, in collaboration with the Technical University of Denmark (DTU). More importantly, we are pursuing a big data approach which scans the PDB for additional structural information and repeats the calculation using this for increased accuracy. We have been validating and applying ZEMu using systems with high potential biomedical impact: Growth Hormone and Staphylococcal protein A. We have submitted our Protein A paper already. A considerable number of our Growth Hormone mutants are now being tested experimentally. Some of our excess time may be used on other projects ongoing in our lab, including a ribosomal frameshifting project in collaboration with Maria Selmer (Uppsala) and Tom Caulfield (Mayo Clinic). A renewed Uppmax CPU time allocation would be very useful to continue this work. Uppmax is preferred as I already have my code compiled and running perfectly there; sometimes a change of system brings new compile/link issues. However if needed I can take an allocation on C3SE or another suitable machine.