Mariusz Milik

Learn More
Proposed is a method for locating functionally relevant atoms in protein structures and a representation of spatial arrangements of these atoms allowing for a flexible description of active sites in proteins. The search method is based on comparison of local structure features of proteins that share a common biochemical function. The method does not depend(More)
We present a workflow that leverages data from chemogenomics based target predictions with Systems Biology databases to better understand off-target related toxicities. By analyzing a set of compounds that share a common toxic phenotype and by comparing the pathways they affect with pathways modulated by nontoxic compounds we are able to establish links(More)
A method for generating a full backbone protein structure from the coordinates of ␣-carbons, is presented. The method extracts information from known protein structures to generate statistical positions for the reconstructed atoms. Tests on a set of proteins structures show the algorithm to be of comparable accuracy to existing procedures. However, the(More)
A simple model of short range interactions is proposed for a reduced lattice representation of polypeptide conformation. The potential is derived on the basis of statistical regularities seen in the known crystal structures of globular proteins. This potential accounts for the generic stiffness of polypeptides, the correlation between peptide bond plates,(More)
A system is presented for generating peptide sequences with desirable properties , using combination of neural network and artificial evolution. The process is illusu·ated by an example of a practical problem of generating at1ificial transbilayer pep tides. The peptides generated in the process of artificial evolution have the physico-chemical properties of(More)
  • 1