Machine Learning for Personalized Medicine

Marie-Curie Action: "Initial Training Networks"

Project: Modeling functional and regulatory modules using statistical and machine learning methods

Cankut Çubuk
Joaquin Dopazo
Valencia, Spain


Cankut Çubuk was born in Denizli, Turkey, in 1985. He completed his undergraduate degree in biochemistry at Ege University, Izmir, Turkey. Afterwards, he worked as an application specialist in a life science company for 3 years. During this time, he started to have interest in high-throughput genomic data analysis. Based on this interest, he started his master degree in Medical Informatics at Dokuz Eylul University, Izmir, Turkey.
In 2012 he completed an internship at Computational Genomics Department, CIPF, Valencia, Spain. In 2013 he participated in a research project of the Department of Medical Technologies and Technomathematik, FH-Aachen, Germany, for 6 months. He has been involved in many different projects from wet laboratory to bioinformatics analysis. In November 2013 he joined the MLPM-Marie Curie Initial Training Network at CIPF as a PhD student.

Project description

Metabolism is a key contributor to major human diseases such as diabetes, cardiovascular disease and cancer. Genetic differences in metabolic pathways can affect individual responses to drugs, both in terms of therapeutic effect as well as adverse effects. A metabolic pathway is series of chemical reactions in a cell, each reaction in a pathway being catalyzed by an enzyme, thus we can assume gene active/inactive status as proxies for enzymes. For our project we are using Genome Scale Human Metabolic Network and high throughput genomic data to improve steady-state flux distributions and by extension, metabolic patterns detection. We are using different statistical and mathematical approaches to obtain a quantitative model of the human metabolic network. Better understanding of tissue, disease and treatment specific regulation of the metabolic pathways will let us to reach the level of P4 (predictive, preventive, personalized and participatory) medical approach.

Motivation for participating in the network

I am participating in the network as a PhD from Prince Felipe Research Center (CIPF), Spain. My current research interest is modeling functional and regulatory modules using statistical and machine learning methods for personalized medicine. MLPM gives me the opportunity to collaborate with several academic labs and private companies that are experts of the same research field. Furthermore, MLPM is a Marie Curie Initial Training Network which combines education, research and training to educate interdisciplinary experts of science. My project as well as my research career will benefit from all given opportunities in this network.


Duration of fellowship: from November 2013 to November 2016



  • Diniz G, Aktas S, Cubuk C, Ortaç R, Vergin C & Olgun N. Tissue expressions of MLH1, PMS2, MSH2, MSH6 proteins and prognostic value of microsatellite instability in wilms tumor: Experience with 45 cases. Pediatric Hematology and Oncology 2013; 30: 273-284. (Doi:10.3109/08880018.2013.780274)

 C. Cubuk’s technical support acknowledged in:

  • Tuzun F, Kumral A, Duman & Ozkan H. Breast Milk Jaundice: The impact of bacteria present in breast milk and infant feces. Journal of Pediatric Gastroenterology & Nutrition 2013; 56: 328–332. (Doi: 10.1097/MPG.0b013e31827a964b)
  • Kumral A, Tuzun F, Tugyan K, Ozbal S, Yılmaz O, Yesilirmak CD, Duman & Ozkan H. Role of epigenetic regulatory mechanisms in neonatal hypoxic-ischemic brain injury. Early Human Development 2013; 89: 165-173. (Doi: 10.1016/j.earlhumdev.2012.09.016.)
  • Kumral A, Tuzun F, Ozbal S, Ugur Ergur B, Yılmaz O, Duman N & Ozkan H (2012). Lipopolysaccharide-preconditioning protects against endotoxin-induced white matter injury in the neonatal rat brain. Brain Research 2012; 1489: 81-89. (Doi: 10.1016/j.brainres.2012.10.015)
  • Kumral A, Tuzun F, Yesilirmak DC, Duman N & Ozkan H. Genetic basis of apnoea of prematurity and caffeine treatment response: role of adenosine receptor polymorphisms: genetic basis of apnoea of prematurity. Acta Paediatrica 2012; 101: 299-303. (Doi: 10.1111/j.1651-2227.2012.02664.x)
  • Aktas S, Zadeogluları Z, Erçetin P & Olgun N. The Effect of Differentiating and Apoptotic Agents on Notch Signalling Pathway in Hepatoblastoma. Hepato-Gastroenterology 2010; 57: 891-897.

MLPM2012 Publications:


- Salavert F, Hidalgo MR, Amadoz A, Cubuk C,  Medina I, Crespo D, Carbonell-Caballero J, Dopazo J. Actionable pathways: interactive discovery of therapeutic targets using signaling pathway models. Nucl. Acids Res. 2016;  doi: 10.1093/nar/gkw369

- Sanchez V, Cubuk C, Sebastián-Leon P, Carobbio S, Dopazo J, Vidal-Puig A, Bartolomucci A. Chronic subordination stress selectively downregulates the insulin signaling pathway in liver and skeletal muscle but not in adipose tissue of male mice. Stress: Int. J. Biol. Stress. 2016; doi: 10.3109/10253890.2016.1151491.

- Razzoli M, Frontini A, Gurney A, Mondini E, Cubuk C, Katz LS, Cero C, Bolan PJ, Dopazo J, Vidal-Puig A, Cinti S, Bartolomucci A. Stress-induced activation of brown adipose tissue prevents obesity in conditions of low adaptive thermogenesis. Molecular Metabolism 2015 (in press).
doi: 10.1016/j.molmet.2015.10.005.

- Alonso R, Salavert F, Garcia-Garcia F, Carbonell-Caballero J, Bleda M, Garcia-Alonso L, Sanchis-Juan A, Perez-Gil D, Marin-Garcia P, Sanchez R, Cubuk C, Hidalgo MR, Amadoz A, Hernansaiz-Ballesteros RD, Aleman A, Tarraga J, Montaner D, Medina I, Dopazo J. Babelomics 5.0: functional interpretation for new generations of genomic data. Nucleic Acids Research 2015; doi: 10.1093/nar/gkv384 


- Cankut Cubuk, Marta R. Hidalgo, Jose Carbonell-Caballero and Joaquín Dopazo. Signalling circuit activities as mechanism-based features to predict mode of action of chemicals. CAMDA 2015 Conference at ISMB/ECCB, Dublin

- Marta R. Hidalgo, Cankut Cubuk, Jose Carbonell-Caballero and Joaquín Dopazo. Functional hallmarks in clear cell renal cell carcinoma grade and stage progression revealed by changes in signalling circuit activities. CAMDA 2015 Conference at ISMB/ECCB, Dublin


Cankut Cubuk, Marta R. Hidalgo, Jose Carbonell-Caballero and Joaquin Dopazo. Identification of Key Metabolic Patterns of Cancer Using RNA-Seq Data. 4th Conference on Constraint-Based Reconstruction and Analysis (COBRA 2015)