Although diseased cells may harbor hundreds of genomic alterations, only a subset of these alterations is driving the disease initiation and progression; these are known as (disease specific) essential genes. Our approach focuses on controlling disease-specific essential genes, as acting upon them is guaranteed to kill the diseased (and only the diseased!) cells. Our new machine learning algorithms identify nodes targetable by FDA-approved drugs, which lead to controlling essential genes, through (sometimes many) cascading effects in the network.
Network pharmacology aims at understanding how and where in the disease network should one target to inhibit disease phenotypes. Since the exponentially increasing number of potential drug combinations makes pure experimental approach quickly unfeasible, there is a major need of computational models and algorithms that can effectively reduce the search space and determine the most promising combinations.
Our approach focuses on identifying sets of drugs with controlling ability over diseases-specific essential genes. Our new machine learning algorithms identifies genes/proteins targetable by FDA-approved drugs, which lead to controlling the expression of essential genes, through cascading effects in the network.
Our approach can analytically predict combinations of 6-10 drugs (or drug-compounds) which are mathematically proved to influence the disease specific essential genes.
Our approach has benefits for three business sectors: pharmaceutical companies, health service companies with strong R&D activities in therapy development and companies offering specialized bio-medical data analysis and personalized medicine.
More information about the project here.
Ion Petre, Professor
+358 (0)2 215 3361
|Vladimir Rogojin||Dwitiya Tiwari|