Processes shaping the 3-dimensional (3D) organization of the human genome remain elusive. A strategy to study genome architecture is to model 3D chromatin structure and analyze properties of structures. However, there is no user-friendly framework enabling high-resolution human genome modeling over a wide range of scales. We introduce Chrom3D, a user-friendly whole-genome 3D modeling platform that enables genome-wide high-resolution modeling of the genome in space, predicting the positioning of topologically-associating domains (TADs) with respect to each other and with respect to the nuclear periphery. Chrom3D integrates chromosome conformation capture (Hi-C) data and lamin-associated domain (LAD) information to generate high-resolution ensembles of structures that recapitulate the radial distribution of TADs detected in single cells. Chrom3D reveals TADs constitutively placed at the nuclear periphery or towards the nuclear interior across structures. TAD stability in these compartments is consistent with their gene density and gene expression level. A- and B-type lamins used as radial constraints differentially skew radial LAD distribution towards the nuclear interior or periphery, respectively. Predictions of radial LAD placement in genome structure ensembles are validated by quantitative imaging. Our structures ascribe a subset of laminopathy-causing lamin A/C mutant LADs in the nuclear interior without any prior knowledge of such localization. Further, Chrom3D structures reveal unexpected features of LAD regulation in the nuclear interior in fibroblasts from laminopathy patients carrying a lipodystrophy-causing LMNA mutation. Integration of radial positioning constraints in 3D genome structures enables the study of spatial gene regulation in health and disease.