Intro PPI alignment Algorithms Results Global alignment of protein-protein interaction networks by graph matching methods. Mikhail Zaslavskiy 1 , Francis Bach 2 and Jean-Philippe Vert 1 1 Mines ParisTech / Institut Curie / INSERM 2 INRIA / Ecole normale superieure de Paris SMPGD 2010. M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Outline 1 Biological context 2 Protein-protein interaction network alignment 3 Algorithms 4 Experimental results 5 Conclusions M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Outline 1 Biological context 2 Protein-protein interaction network alignment 3 Algorithms 4 Experimental results 5 Conclusions M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Proteins M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Functional orthologs Species 1 Species 2 f 1 : MKQALAAADDDDAQ... y 1 : MDDDDALGLLLLA... f 2 :MGDXLLMMAALLLL... y 2 : MHHAAKLLDDAS... ... ... Objective : Automatic identification of protein functional orthologs M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Functional orthologs Species 1 Species 2 f 1 : MKQALAAADDDDAQ... y 1 : MDDDDALGLLLLA... f 2 :MGDXLLMMAALLLL... y 2 : MHHAAKLLDDAS... ... ... Objective : Automatic identification of protein functional orthologs M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Sequence comparison: best-best blast hits Species 1 Species 2 f 1 : MKQDLARIEQFLDALF... y 1 : MSRLPVLLLLQLLVRGA. . . f 2 : MSKLKIAVSDSCPDCF... y 2 : MELAALCRAGLLLALDA. . . ... ... y 1 y 2 C = 10 50 C ij -BLAST similarity scores f 1 f 2 27 10 Optimal assignment : f 1 → y 2 , f 2 → y 1 Potential difficulties: y is the best hit for f , but f is not the best hit for y ( y 1 , f ) and ( y 2 , f ) produce very close blast scores. Which one to choose ? M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Sequence comparison: best-best blast hits Species 1 Species 2 f 1 : MKQDLARIEQFLDALF... y 1 : MSRLPVLLLLQLLVRGA. . . f 2 : MSKLKIAVSDSCPDCF... y 2 : MELAALCRAGLLLALDA. . . ... ... y 1 y 2 C = 10 50 C ij -BLAST similarity scores f 1 f 2 27 10 Optimal assignment : f 1 → y 2 , f 2 → y 1 Potential difficulties: y is the best hit for f , but f is not the best hit for y ( y 1 , f ) and ( y 2 , f ) produce very close blast scores. Which one to choose ? M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Sequence comparison: best-best blast hits Species 1 Species 2 f 1 : MKQDLARIEQFLDALF... y 1 : MSRLPVLLLLQLLVRGA. . . f 2 : MSKLKIAVSDSCPDCF... y 2 : MELAALCRAGLLLALDA. . . ... ... y 1 y 2 C = 10 50 C ij -BLAST similarity scores f 1 f 2 27 10 Optimal assignment : f 1 → y 2 , f 2 → y 1 Potential difficulties: y is the best hit for f , but f is not the best hit for y ( y 1 , f ) and ( y 2 , f ) produce very close blast scores. Which one to choose ? M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Speciation M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Speciation M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Speciation & Duplication M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Speciation & Duplication M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Ambiguous assignment resolving Increase the quality of similarity score, 3D structures Protein interactions . . . M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Ambiguous assignment resolving Increase the quality of similarity score, 3D structures Protein interactions . . . M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Outline 1 Biological context 2 Protein-protein interaction network alignment 3 Algorithms 4 Experimental results 5 Conclusions M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Protein interaction networks M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Protein interaction network Fly interaction network M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Ambiguous assignment resolving M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Ambiguous assignment resolving M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Ambiguous assignment resolving M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Ambiguous assignment resolving M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Ambiguous assignment resolving M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Ambiguous assignment resolving ⇒ Maximize :the number of overlapping edges M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Ambiguous assignment resolving ⇒ Maximize :the number of overlapping edges M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Two types of global network alignment Constrained Alignment: Only proteins having close sequence structures may be matched to each other InParanoid clusters: Maximize : J = the number of overlapping edges Balanced Alignment: Let S ij = BLAST ( f i , y j ), then the total sequence similarity � S = S ij f i is matched to y j Maximize : (1 − λ ) J + λ S λ controls the trade-off between J and S . M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Two types of global network alignment Constrained Alignment: Only proteins having close sequence structures may be matched to each other InParanoid clusters: Maximize : J = the number of overlapping edges Balanced Alignment: Let S ij = BLAST ( f i , y j ), then the total sequence similarity � S = S ij f i is matched to y j Maximize : (1 − λ ) J + λ S λ controls the trade-off between J and S . M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Two types of global network alignment Constrained Alignment: Only proteins having close sequence structures may be matched to each other InParanoid clusters: Maximize : J = the number of overlapping edges Balanced Alignment: Let S ij = BLAST ( f i , y j ), then the total sequence similarity � S = S ij f i is matched to y j Maximize : (1 − λ ) J + λ S λ controls the trade-off between J and S . M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Two types of global network alignment Constrained Alignment: S. Bandyopadhyay, R. Sharan and T. Ideker, 2006 : MRF model, approximate solution Our contribution: exact solution by the Message Passing Algorithm Balanced Alignment: S. R. Singh, J. Xu and B. Berger, 2008 : IsoRank algorithm Our contribution: Graph Matching Algorithms M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Two types of global network alignment Constrained Alignment: S. Bandyopadhyay, R. Sharan and T. Ideker, 2006 : MRF model, approximate solution Our contribution: exact solution by the Message Passing Algorithm Balanced Alignment: S. R. Singh, J. Xu and B. Berger, 2008 : IsoRank algorithm Our contribution: Graph Matching Algorithms M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Outline 1 Biological context 2 Protein-protein interaction network alignment 3 Algorithms 4 Experimental results 5 Conclusions M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Constrained Alignment: Message Passing MP is based on a forward-backward recursion similar to the Viterbi recursion for hmm. M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Constrained Alignment: Message Passing Forward step: compute the number of conserved interactions starting from leaves M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Intro PPI alignment Algorithms Results Constrained Alignment: Message Passing Forward step: compute the number of conserved interactions starting from leaves M. Zaslavskiy, F. Bach, JP. Vert Global alignment of PPI networks
Recommend
More recommend