Discovering putative prion sequences in complete proteomes using probabilistic representations of Q/N-rich domains | Publicación