G-quadruplexes (G4) are non-canonical structures formed by guanine-rich sequences of nucleic acids that have been shown to exist in living cells where they are thought to participate in regulation of gene expression. DNA G-quadruplexes are highly polymorphic and can fold into a variety of three-dimensional structures. An often delicate conformational equilibrium between different possible G4 structures depends on a variety of different factors, in a manner not yet fully understood. The goal of the current project is to advance our understanding of this sequence-structure relation so as to allow for a comprehensive model-based design of G4s with a desired folded conformation. This will involve determination of conformational landscapes of a wide range of G4-forming sequences and analysis of these landscapes for structural and energetic principles underlying the relative stability of possible G4 folds. The project focuses on three primary research objectives that will be addressed using an integrative approach combining multiscale modeling and complementary experimental methods. 1) Determination and validation of the conformational stability maps for an extensive set of G-rich DNA sequences prone to form two-tetrad G-quadruplexes. We will first use molecular dynamics (MD) simulations to evaluate relative stabilities of all theoretically possible two-tetrad G4 conformations for a systematic set of G4-forming sequences. Next, these predictions will be validated using an array of experimental methods, including 1H NMR, CD and UV spectroscopy. 2) Detailed analysis of structural and energetic factors underlying the relative stability of possible G4 folded states. To explain observed conformational behavior, we will analyze the above obtained data sets by means of statistical methods. The computed folding free energies will be decomposed into enthalpic and entropic contributions, allowing for understanding the molecular basis for stability of G4s. 3) Extending the sequence-structure analysis to three-tetrad G-quadruplexes and testing the model as a tool for prediction and design of G4 conformations. To evaluate and, if necessary, extend the applicability of our treatment to the three-tetrad G-quadruplexes we will test two major hypotheses about the energetic relation between the three- and two-tetrad G4 structures. Finally, we will assess the predictive power of our model by designing and experimentally validating a set of three-tetrad G4 conformations.