Remodeling of nuclear organization occurs during normal cell development, differentiation and cancer. One of the biggest gaps of knowledge remains how to link the information on chromatin and chromosome structural organization with genes activity. In this paper we introduce some physical ideas and a general computational method demonstrating how genome 3D architecture and its remodeling can be quantitatively modeled. We study a hypothetical scenario of alterations of chromosome territories positioning in the course of cell proliferation. On this basis we obtain quantitative information about chromosomal contacts in the nucleus. We predict changes of radial distributions of contacts between chromosomal megabase domains during proliferation. The proposed modeling approach may be helpful in integrating experimental data on nuclear reorganization associated with normal development and with various diseases. This predictive modeling may find applications in genome research of normal and cancer cells, stem cell biology, biology of aging, etc.