Topological autocorrelations for prediction of protein conformational stability and kinase and protease inhibitions トポロジカル自己相関関数を用いた蛋白質の構造安定性とキナーゼ及びプロテアーゼ阻害の予測

この論文をさがす

著者

    • Fernandez Llamosa Michael フェルナンデス ヤモサ ミチャエル

書誌事項

タイトル

Topological autocorrelations for prediction of protein conformational stability and kinase and protease inhibitions

タイトル別名

トポロジカル自己相関関数を用いた蛋白質の構造安定性とキナーゼ及びプロテアーゼ阻害の予測

著者名

Fernandez Llamosa Michael

著者別名

フェルナンデス ヤモサ ミチャエル

学位授与大学

九州工業大学

取得学位

博士(情報工学)

学位授与番号

情工甲第250号

学位授与年月日

2011-03-25

注記・抄録

博士論文

九州工業大学博士学位論文(要旨) 学位記番号:情工博甲第250号 学位授与年月日:平成23年3月25日

The annotation of protein structure and function from sequence and the prediction of compound’s activity from sketch representations are fundamental goals in bio- and chemoinformatics. In the present study, fast and accurate predictors of protein conformational stability and kinase and protease inhibitions were built from graph representations of proteins and ligands. Firstly, Amino Acid Sequence Autocorrelation (AASA) vectors were computed from the Cα-carbon linear graph representation of a large dataset of protein mutants (>1000) from Protherm database.Genetic algorithm-optimized support vector machines (GA-SVM) were trained with AASA vectors to predict the real ΔΔG values with squared correlation coefficient of 0.45 and classify ΔΔG signs with accuracy of 80%. The stable mutants in the test set were recognized with accuracies of 70%. Secondly, AASA vectors and ligand’s autocorrelation features were computed from the linear graph representation of kinase and protease and from 2D molecular graphs collected from ProLINT database. SVMs trained with concatenated autocorrelation matrices yielded test set accuracies > 80% for kinase and protease targets. The inhibition predictors perform homogenously along the different kinase and protease families and ligands’ scaffolds. The predictors from sequences and sketch representations of ligands are online available at:http://gibk21.bse.kyutech.ac.jp/llamosa/ddG-AASA/ddG_AASA.html http://gibk21.bse.kyutech.ac.jp/AUTOkinI/SVMpredictor.html http://gibk21.bse.kyutech.ac.jp/AUTOprotI/SVMpredictor.html

九州工業大学博士学位論文 学位記番号:情工博甲第250号 学位授与年月日:平成23年3月25日

CHAPTER 1.INTRODUCTION|CHAPTER 2.DATASETS AND COMPUTATIONAL METHODS|CHAPTER 3.MODELING OF PROTEIN CONFORMATIONAL STABILITY|CHAPTER 4.MODELING OF KINASE INHIBITION|CHAPTER 5.MODELING OF PROTEASE INHIBITION|CHAPTER 6. SUMMARY AND FUTURE PERSPECTIVES

平成22年度

1アクセス

各種コード

  • NII論文ID(NAID)
    500000570369
  • NII著者ID(NRID)
    • 8000000572656
  • 本文言語コード
    • eng
  • NDL書誌ID
    • 024444966
  • データ提供元
    • 機関リポジトリ
    • NDL ONLINE
ページトップへ