TY - GEN
T1 - Inferring conict-sensitive phosphorylation dynamics
AU - Cheng, Qiong
AU - Ogihara, Mitsunori
AU - Gupta, Vineet
PY - 2011
Y1 - 2011
N2 - Phosphorylation is a ubiquitous and fundamental regulatory mechanism that controls signal transduction in living cells. It acts as switches in cell communications. One of the most important factors contributing to the dynamics is the binding competition, which refers to the existence of more than one protein that physically binds to the same or an overlapping residue on a protein. By integrating data from different sources on the HeLa cancer cells and all available Homo sapiens (human) cells and iteratively examining the phosphorylation interfaces, we found a number of conicting interaction pairs. We extended the search into indirect conicts over direct upstream cascades and further into the whole network and calculated a min-max conict-sensitive decomposition of phosphorylation network by graph-theoretical methods. Further we used EGF-stimulation phosphoproteome data and obtained activation patterns of phosphorated proteins by soft clustering. By combining these two groupings, we calculated an optimal conict-free activation patterns using maximum bipartite matching. Sorting the average peak time of the activation patterns brought forth an activation order of min-max conict-sensitive decomposition subnetworks. We evaluated conict-sensitive phosophorylation dynamics by analyzing the importance of the conicting interactions in the whole networks, the distribution of serine/threonine/tyrosine phosphorylation, and the direct or indirect activation order of phosphorylated proteins. Compared with a previously published approach [15], our solution discovered conict-sensitive dynamics that resolved conicts and it inferred more practical causal effects consistent with EGFR signaling pathways.
AB - Phosphorylation is a ubiquitous and fundamental regulatory mechanism that controls signal transduction in living cells. It acts as switches in cell communications. One of the most important factors contributing to the dynamics is the binding competition, which refers to the existence of more than one protein that physically binds to the same or an overlapping residue on a protein. By integrating data from different sources on the HeLa cancer cells and all available Homo sapiens (human) cells and iteratively examining the phosphorylation interfaces, we found a number of conicting interaction pairs. We extended the search into indirect conicts over direct upstream cascades and further into the whole network and calculated a min-max conict-sensitive decomposition of phosphorylation network by graph-theoretical methods. Further we used EGF-stimulation phosphoproteome data and obtained activation patterns of phosphorated proteins by soft clustering. By combining these two groupings, we calculated an optimal conict-free activation patterns using maximum bipartite matching. Sorting the average peak time of the activation patterns brought forth an activation order of min-max conict-sensitive decomposition subnetworks. We evaluated conict-sensitive phosophorylation dynamics by analyzing the importance of the conicting interactions in the whole networks, the distribution of serine/threonine/tyrosine phosphorylation, and the direct or indirect activation order of phosphorylated proteins. Compared with a previously published approach [15], our solution discovered conict-sensitive dynamics that resolved conicts and it inferred more practical causal effects consistent with EGFR signaling pathways.
UR - http://www.scopus.com/inward/record.url?scp=84858970769&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84858970769&partnerID=8YFLogxK
U2 - 10.1145/2147805.2147864
DO - 10.1145/2147805.2147864
M3 - Conference contribution
AN - SCOPUS:84858970769
SN - 9781450307963
T3 - 2011 ACM Conference on Bioinformatics, Computational Biology and Biomedicine, BCB 2011
SP - 430
EP - 434
BT - 2011 ACM Conference on Bioinformatics, Computational Biology and Biomedicine, BCB 2011
T2 - 2011 ACM Conference on Bioinformatics, Computational Biology and Biomedicine, ACM-BCB 2011
Y2 - 1 August 2011 through 3 August 2011
ER -