Robert Gray Dodge Chair Installation Ceremony: Laszlo Barabasi

Northeastern researchers connect diseases based on their molecular similarities

North­eastern Uni­ver­sity net­work sci­en­tists have found a way to con­nect dis­eases based on their shared mol­e­c­ular inter­ac­tions. Pub­lished in the journal Sci­ence, the North­eastern team cre­ated a math­e­mat­ical tool to ana­lyze the human interactome—a map of the mol­e­c­ular inter­ac­tions within cells—and found that over­lap­ping dis­ease modules—neighborhoods of disease-​​associated proteins—result in some­times unex­pected rela­tion­ships between diseases.

The find­ings con­sti­tute a remark­able step in under­standing human dis­eases. “It is increas­ingly obvious that human dis­eases can be inter­preted only in the con­text of the intri­cate mol­e­c­ular net­work between the cell’s com­po­nents,” said Albert-​​László Barabási, Robert Gray Dodge Pro­fessor of Net­work Sci­ence and Uni­ver­sity Dis­tin­guished Pro­fessor and director of Northeastern’s Center for Com­plex Net­work Research. “What was not obvious until now is whether the avail­able net­work maps offer enough cov­erage and accu­racy to help us get started in this path. In this paper we showed that they do offer that accu­racy and pro­vide valu­able infor­ma­tion about the mol­e­c­ular ori­gins of disease-​​disease relationships.”

The team ana­lyzed 299 dis­eases that had at least 20 asso­ci­ated genes and found that 226 of the dis­eases had their own spe­cific “neigh­bor­hood” within the inter­ac­tome. They also dis­cov­ered that dis­eases that were far away from each other within the inter­ac­tome had very little in common in terms of mol­e­c­ular func­tions or symp­toms, while ones in the same “neigh­bor­hood” were more similar.

Shared genes offer only lim­ited infor­ma­tion about the rela­tion­ship between two dis­eases. By applying their net­work sci­ence tools to ana­lyze the inter­ac­tome, Barabasi and his team found that two seem­ingly unre­lated dis­eases can actu­ally be con­nected based on the net­work dis­tance between the dis­ease modules. For example, they found that asthma, a res­pi­ra­tory dis­ease, and celiac dis­ease, an autoim­mune dis­ease of the small intes­tine, are local­ized in over­lap­ping neigh­bor­hoods sug­gesting shared mol­e­c­ular roots despite their rather dif­ferent pathobiologies.

What we try to achieve in this paper is to lay out a more math­e­mat­ical ground­work for this intu­itive idea that you can use the inter­ac­tome as a map,” said Jörg Menche, a postodoc­toral researcher and one of the authors on the paper. “It could play a major role in under­standing dis­eases on a mol­e­c­ular lev­e­land devel­oping better remedies.”

Menche offered an example of the power of the net­work map, explaining that doc­tors typ­i­cally diag­nose patients based on symp­toms described by the patient. But using the net­work map, he said, could make it pos­sible to find out what is hap­pening in the gene product inter­ac­tions to cause the particular ailment.

Despite impres­sive advances in inter­ac­tome map­ping and dis­ease gene iden­ti­fi­ca­tion, both the inter­ac­tome and our knowl­edge of disease-​​associated genes are still woe­fully incom­plete. This incom­plete­ness prompted us to sys­tem­at­i­cally inves­ti­gate to what extent the cur­rent data are suf­fi­cient to map out dis­ease mod­ules,” added Mensche.

In con­clu­sion,” he said “we could show that the cur­rently avail­able inter­ac­tome data has reached suf­fi­cient cov­erage to sys­tem­at­i­cally inves­ti­gate the mol­e­c­ular mech­a­nisms under­lying many dis­eases, as well as to explore patho­bi­o­log­ical rela­tion­ships between dis­eases on a mol­e­c­ular level.”

The North­eastern researchers are based in the Center for Com­plex Net­work Research. The team com­prises Barabási, Menche, post­doc­toral researcher Maskim Kitsak, research assis­tant pro­fessor Amitabh Sharma, and grad­uate physics stu­dent Susan Dina Ghiassian, PhD’15.

Originally published in news@Northeastern on February 23, 2015.

College of Science