MANILA, Philippines (May 2026) — Finding a reliable drug target in a sea of competing biological models has always been one of science’s messiest problems. A team of Filipino mathematicians may have just made it a lot cleaner.
Researchers from the University of the Philippines Diliman and De La Salle University developed a new analytical method called Common Species Embedded Networks (CSEN) analysis, which allows scientists to meaningfully compare different mathematical models of the same biological process — something that has historically been difficult to do.
The problem with too many models
Biological processes like the Wnt signaling pathway, which plays a critical role in cell development and tissue maintenance, are often mapped out by scientists as reaction networks. These are essentially mathematical blueprints that describe how proteins and chemicals interact inside the body. The catch: different research groups often produce different versions of these maps for the same process, and comparing them has been like trying to read two books written in slightly different languages.
Dr. Bryan Hernandez of the UP Diliman College of Science’s Institute of Mathematics, together with Patrick Vincent Lubenia and Dr. Eduardo Mendoza of DLSU’s Center for Natural Science and Environmental Research, built CSEN analysis to solve exactly that.
How it works
The method zeroes in on the species — proteins, chemicals, and other biological components — that two models share. It then checks whether the reactions involving those shared components are equivalent, or if mathematical transformations can bridge the differences between them.
“The method works by first isolating the networks ’embedded’ within the models that involve only the common species,” Dr. Hernandez explained. “For the reactions that are not identical, the method checks for ‘transformations’ — mathematical links that can explain how one reaction set might induce equivalence between the systems.”
When the team applied CSEN to several existing Wnt signaling models, including those by Lee, Schmitz, MacLean, and Feinberg, they found that some models previously thought to be fundamentally different are actually structurally very similar. In other cases, the analysis confirmed that no meaningful relationship existed between models at all.
Why it matters for medicine
The practical payoff is significant. When multiple models — despite their differences — all point to the same biological interaction as a driver of disease, that interaction becomes a more credible and reliable therapeutic target.
CSEN also helps researchers identify which parts of a model are redundant and which are genuinely unique, allowing them to refine and consolidate competing theories rather than treating each one as a separate entity.
“Historically, it has been difficult to compare these models because they are often treated as entirely separate entities, utilizing different sets of variables and reactions,” Dr. Hernandez noted.
Beyond biology
The researchers say CSEN is not limited to Wnt signaling or even biology in general. The same framework could, in theory, be applied to insulin signaling, metabolic pathways, chemical engineering processes, or ecological models — any system that can be represented as a reaction network.
The study, titled “Embedding-based comparison of reaction networks of Wnt signaling,” was published in MATCH Communications in Mathematical and in Computer Chemistry, an open-access journal covering mathematical applications in chemistry.
