Most of the momentum of a proton is carried by the quarks I saw this in wiki
https://en.wikipedia.org/wiki/Antiproton#Modern_experiments_and_applications
"the valence quarks in the proton, and the valence antiquarks in the antiproton, tend to carry the largest fraction of the proton or antiproton's momentum."
However valence quarks only account for a very small fraction of the proton's mass so how do they carry most of its momentum? Or is it virtual particles that does that?
 A: Momentum, not mass. The sentence you are misreading refers to highly energetic protons with kinetic energies dozens of times higher than the mass of the nucleon or antinucleon in question, the realm of perturbative QCD. The valence partons involved carry a fraction x of the momentum of the struck nucleon —see the formal definition in the PDG review — of the order of 1/3, as visible in the green and blue distributions of the WP article figure. By contrast, "sea" quarks carry smaller fractions of the momentum. Gluons carry collectively about a half.
These are completely different fractions and entities than the accounting of operators expectations fashionable in computing nucleon masses in highly nonperturbative QCD via lattice simulations, and reported rather ineptly in the popular science press: veritable recipes for misconceptions. The quark operators in all of these accountings are definitely not valence quarks, they are ultra-light current (fundamental) quarks, and the small fraction is an operator corresponding to just their mass terms ("condensate"), not their over-all  "energy" and "anomaly" contribution, the bulk of the "heft". 
The methods, culture, significance and mental pictures involved in each of the above paragraphs are so disparate that calling them apples and oranges is flattering them (!) with conceptual connections that simply are not there.  A particle physics student is warned early and often to never put these two domains in his visual field or mental buffer simultaneously, as ineffable grief cannot fail to ensue...
