“Strat­egy changes are an essen­tial part of evo­lu­tion­ary games. Here we intro­duce a sim­ple rule that, depend­ing on the value of a sin­gle para­me­ter $w$, influ­ences the selec­tion of play­ers that are con­sid­ered as poten­tial sources of the new strat­egy. For pos­i­tive $w$ play­ers with high pay­offs will be con­sid­ered more likely, while for neg­a­tive $w$ the oppo­site holds. Set­ting $w$ equal to zero returns the fre­quently adopted ran­dom selec­tion of the oppo­nent. We find that increas­ing the prob­a­bil­ity of adopt­ing the strat­egy from the fittest player within reach, i.e. set­ting $w$ pos­i­tive, pro­motes the evo­lu­tion of coop­er­a­tion. The robust­ness of this obser­va­tion is tested against dif­fer­ent lev­els of uncer­tainty in the strat­egy adop­tion process and for dif­fer­ent inter­ac­tion net­work. Since the evo­lu­tion to wide­spread defec­tion is tightly asso­ci­ated with coop­er­a­tors hav­ing a lower fit­ness than defec­tors, the fact that pos­i­tive val­ues of $w$ facil­i­tate coop­er­a­tion is quite surprising. …”

“In this work we study a weak Prisoner’s Dilemma game in which both strate­gies and update rules are sub­jected to evo­lu­tion­ary pres­sure. Inter­ac­tions among agents are spec­i­fied by com­plex topolo­gies, and we con­sider both homo­ge­neous and het­ero­ge­neous sit­u­a­tions. We con­sider deter­min­is­tic and sto­chas­tic update rules for the strate­gies, which in turn may con­sider sin­gle links or full con­text when select­ing agents to copy from. Our results indi­cate that the co-​​evolutionary process pre­serves het­ero­ge­neous net­works as a suit­able frame­work for the emer­gence of coop­er­a­tion. Fur­ther­more, on those net­works, the update rule lead­ing to a larger frac­tion of coop­er­a­tion, repli­ca­tor dynam­ics, is selected dur­ing co-evolution.…We con­clude that for a vari­ety of topolo­gies, the fact that the dynam­ics coe­volves with the strate­gies leads in gen­eral to more coop­er­a­tion in the weak Prisoner’s Dilemma game.”

“Dynam­ics of evo­lu­tion­ary games strongly depend on under­ly­ing net­works. We study the coevo­lu­tion­ary prisoner’s dilemma in which play­ers change their local net­works as well as strate­gies (i.e., coop­er­ate or defect). This topic has been increas­ingly explored by many researchers. On the basis of active link­ing dynam­ics [J. M. Pacheco et al., J. Theor. Biol. 243, 437 (2006), J. M. Pacheco et al., Phys. Rev. Lett. 97, 258103 (2006)], we show that coop­er­a­tion is enhanced fairly robustly. In par­tic­u­lar, coop­er­a­tion evolves when the pay­off of the player is nor­mal­ized by the num­ber of neigh­bors; this is not the case in the evo­lu­tion­ary prisoner’s dilemma on sta­tic networks.”

“Bio­log­i­cal net­works of inter­act­ing agents exhibit sim­i­lar topo­log­i­cal prop­er­ties for a wide range of scales, from cel­lu­lar to eco­log­i­cal lev­els, sug­gest­ing the exis­tence of a com­mon evo­lu­tion­ary ori­gin. A gen­eral evo­lu­tion­ary mech­a­nism based on global sta­bil­ity has been pro­posed recently [J I Per­otti, O V Bil­loni, F A Tamarit, D R Chialvo, S A Can­nas, Phys. Rev. Lett. 103, 108701 (2009)]. This mech­a­nism is incor­po­rated into a model of a grow­ing net­work of inter­act­ing agents in which each new agent’s mem­ber­ship in the net­work is deter­mined by the agent’s effect on the network’s global sta­bil­ity. We show that, out of this sta­bil­ity con­straint, sev­eral topo­log­i­cal prop­er­ties observed in bio­log­i­cal net­works emerge in a self orga­nized man­ner. The influ­ence of the sta­bil­ity selec­tion mech­a­nism on the dynam­ics asso­ci­ated to the result­ing net­work is ana­lyzed as well.”

“How do liv­ing cells achieve suf­fi­cient abun­dances of func­tional pro­tein com­plexes while min­i­miz­ing promis­cu­ous non-​​functional inter­ac­tions between their pro­teins? Here we study this prob­lem using a first-​​principle model of the cell whose phe­no­typic traits are directly deter­mined from its genome through bio­phys­i­cal prop­er­ties of pro­tein struc­tures and bind­ing inter­ac­tions in crowded cel­lu­lar envi­ron­ment. The model cell includes three inde­pen­dent path­ways, whose topolo­gies of PPI sub­net­works are dif­fer­ent, but whose func­tional con­cen­tra­tions equally con­tribute to cell’s fit­ness. The model cells evolve through geno­typic muta­tions and phe­no­typic pro­tein copy num­ber vari­a­tions. We found a strong rela­tion­ship between evolved physical-​​chemical prop­er­ties of pro­tein inter­ac­tions and their abun­dances due to a “frus­tra­tion” effect: strength­en­ing of func­tional inter­ac­tions brings about hydropho­bic sur­faces, which make pro­teins prone to promis­cu­ous binding.…”

“We intro­duce the het­ero­ge­neous voter model (HVM), in which each agent has its own intrin­sic rate to change state, reflec­tive of the het­ero­gene­ity of real peo­ple, and the par­ti­san voter model (PVM), in which each agent has an innate and fixed pref­er­ence for one of two pos­si­ble opin­ion states. For the HVM, the time until con­sen­sus is reached is much longer than in the clas­sic voter model. For the PVM in the mean-​​field limit, a pop­u­la­tion evolves to a “self­ish” state, where each agent tends to be aligned with its inter­nal pref­er­ence. For finite pop­u­la­tions, dis­crete fluc­tu­a­tions ulti­mately lead to con­sen­sus being reached in a time that scales expo­nen­tially with pop­u­la­tion size.”