North­eastern Uni­ver­sity researchers are offering a fas­ci­nating glimpse into how greater con­trol of com­plex sys­tems, such as cel­lular net­works and social media, can be achieved by merging the tools of net­work sci­ence and con­trol theory.

Albert-​​László Barabási and Yang-​​Yu Liu coau­thored a paper on the research find­ings, fea­tured as the cover story in the May 12 issue of the journal Nature. Barabási, a world-​​renowned net­work sci­en­tist, is a dis­tin­guished pro­fessor in the Depart­ments of Physics and Biology and the Col­lege of Com­puter and Infor­ma­tion Sci­ence, and is the founding director of Northeastern’s Center for Com­plex Net­work Research. Liu is a post­doc­toral research asso­ciate in Barabasi’s lab.

The researchers said this approach can lead to major strides in under­standing com­plex engi­neering and bio­log­ical sys­tems. For example, con­trol­ling the neural and meta­bolic path­ways in living organ­isms could lead to health-​​care break­throughs in drug dis­covery and dis­ease treatments.

Most large com­plex net­works have been cre­ated for some prac­tical pur­pose: meta­bolic net­works to process the food we eat, the Internet to transfer infor­ma­tion, orga­ni­za­tional net­works to achieve the goals of an orga­ni­za­tion,” said Barabási. “The tools devel­oped in this paper offer the pos­si­bility to better under­stand how to con­trol these sys­tems. This could poten­tially gen­erate more effi­cient meta­bolic path­ways, with appli­ca­tions in devel­oping cures to meta­bolic dis­eases, to offering new insights into the design of better organizations.”

Barabási and Liu col­lab­o­rated with MIT researcher Jean-​​Jacques Slo­tine on the paper.

The researchers note that con­trol theory already offers math­e­mat­ical tools for steering engi­neered and nat­ural sys­tems — such as syn­chro­nized man­u­fac­turing processes, cars, robots and elec­trical cir­cuits — toward a desired state.

How­ever, they said a frame­work is lacking to take charge of com­plex, self-​​organized sys­tems — such as cel­lular and social net­works. To meet this chal­lenge, they com­bined the prin­ci­ples of con­trol theory with their inno­v­a­tive net­work sci­ence research to develop an algo­rithm that can assess the driver nodes, or con­nec­tion points, within a par­tic­ular com­plex net­work. By doing so, they can deter­mine how many nodes are nec­es­sary to con­trol in order to gain con­trol of the system.

The trio was inter­ested in dis­cov­ering the min­imum number of driver nodes needed to con­trol a com­plex net­work. They found that denser net­works with more con­nec­tions — such as online social net­works — were easier to con­trol than cel­lular net­works. They also found that sparse net­works, like many bio­log­ical and com­mu­ni­ca­tion net­works, are the hardest to control.

Liu said this work rep­re­sents a fun­da­mental con­tri­bu­tion to both con­trol theory and net­work sci­ence research.

This work was not pos­sible 10 years ago, because at that time we didn’t know how to cat­e­go­rize these com­plex net­works. We didn’t have the data,” Liu said. “But today, we have the data avail­able for empir­ical studies on many large-​​scale networks.”

View selected pub­li­ca­tions of Albert-​​László Barabási in IRis, Northeastern’s dig­ital archive.