Rosetta is one of the currently most successful AbInitio methods developed by the group of DavidBaker
In the Rosetta method, the chain of a protein is broken into three or nine residue segments. A search in the fragment library was carried out to obtain the distribution of the conformations for each target segment. Then a conformation space was built from the insertion of the segments. New improved Rosetta used a sequence and secondary structure profile-profile camparsion method to do the searching from the fragment library.
And then a MonteCarloMethod - SimulatedAnnealing procedure was used to select favorable structures from the conformation space with an improved energy function. Since lots of decoy structures were generated during the process, a final filtering and [Clustering] procedure was used to select the candidates from the decoys.
You can contribute the calculation via RosettaAtHome.
철자검색기는 제대로 동작하는건가요? 그냥 막 쳐서 틀린철자 많을지도 모르는데 ;; 고쳐주세요~ --[wsjoung]
Rosetta 관련 논문들
Toward High-Resolution de Novo Structure Prediction for Small Proteins
- Science 16 September 2005: Vol. 309. no. 5742, pp. 1868 - 1871
Toward High-Resolution de Novo Structure Prediction for Small Proteins
Rosetta 가 사용하는 low-resolution 과 high-resolution 을 조합한 구조 예측 방법이 소개 되어 있다.