Abstract: An important problem in biology informatics, that is, protein structure prediction, is studied in this paper. Enlightened by the law of reciprocity among things in the physics world, a continuous model in 2D Euclidean space is given which gain an advantage over the discretion model in some aspects. The continuous model’s strong point is probably that it utilizes an external “inartificial leading” which is the gravitation among hydrophobe amino acid. Accordingly, in the presupposition of the equivalent result, the continuous model redounds to improving the computing rapidity. Then, a related quasiphysical algorithm is formulated according to the 2D continuous model. Finally some experiment results are given, which also prove the advantages of the continuous model and its quasiphysical algorithm.