GAGRA is an INFN national specific initiative whose aim is, among other ones, to employ new methods based on recent developments (M. Bochicchio [1,2,3]) for the understanding of confining asymptotically free gauge theories -- specifically of the 't Hooft large-N limit of QCD -- and for the consistent quantization of gravity (D. Anselmi [4,5,6]) that do not rely either on the simplifications furnished by the supersymmetry or on the superstring unification of all the interactions respectively. 

A key role in the development of the aforementioned methods is played by the high-energy features of both large-N QCD (M. Bochicchio [1,3]) or, more generally, asymptotically free gauge theories, and of gravity (D. Anselmi [4,5]), mostly in relation to the non-perturbative (M. Bochicchio [1,3], M. Papinutto [7], D. Anselmi [4,6], F. Scardino) and perturbative (M. Becchetti [8], A. Quadri [9]) renormalization properties of the S-matrix amplitudes and correlators. 

On the gauge side, the aforementioned methods provide the strongest constraints [1,2] on the high-energy non-perturbative structure of large-N QCD, and may be a powerful guide [3] to eventually construct candidates for its partial solution. 

On the gravity side, the aforementioned methods have led to the construction [4] of a renormalizable [5] unitary S matrix for gravity, by means of an underlying quantization of general relativity that violates [6] microcausality.

Moreover, complementary -- but crucial -- aims are the exploration of the non-perturbative low-energy features of QCD by lattice gauge theories (M. Papinutto, F. Russo) and QCD phenomenology (A. Polosa [10], A. Pilloni [10]) and, on the gravity side, the experimental search for dark matter (A. Polosa [11]).