Diego A. Galeano G.

I am a Postdoctoral Researcher in Data Science at Fundação Getulio Vargas (FGV) Rio de Janeiro, Brasil, working with Professor Alberto Paccanaro.

I develop machine learning models for applications in medicine and drug discovery.

I started a postdoc at FGV-Emap Rio de Janeiro, Brasil in February 2020 ;)!

Recent news
  • July 2020 Our paper on Predicting the Frequencies of drug Side effects has been accepted at Nature Communications! This includes an extension on my bioarXiv paper that includes pharmacological interpretation of the model done in collaboration with Dr. Shantao Li and Prof. Mark Gerstein from Yale University.
  • October 2019 Our paper on Learning interpretable disease self-representations for drug repositioning has been accepted at the NeurIPS Graph Representation Learning Workshop!
2019
Predicting the Frequencies of Drug Side Effects
Diego A. Galeano, and Alberto Paccanaro
Preprint 2019.
(bioarxiv)
Learning Interpretable Disease Self-Representations for Drug Repositioning
(co-first author*) Fabrizio Frasca*, Diego A. Galeano*, Guadalupe Gonzalez, Ivan Laponogov, Kirill Veselkov, Alberto Paccanaro, Michael M Bronstein
NeurIPS Graph Representation Learning Workshop. 2019.
pdf (arxiv)
2018
A recommender system approach for predicting drug side effects
Diego A. Galeano, and Alberto Paccanaro
IJCNN. 2018.
(IEEE Xplore)
2017
Mining the Biomedical Literature to predict shared drug targets in drugbank
(co-first author*) Horacio Caniza*, Diego A. Galeano*, and Alberto Paccanaro
IEEE-CLEI. 2017.
(IEEE Xplore)
2016
Drug targets prediction using chemical similarity
Diego A. Galeano, and Alberto Paccanaro
IEEE-CLEI. 2016.
2015
Posturography Platform and Balance Control Training and Research System Based on FES and Muscle Synergies
Diego A. Galeano, Fernando Brunetti, Diego Torricelli, S. Piazza, JLP Rovira
Springer Series in Computational Neuroscience (book). 2015.
2014
A tool for balance control training using muscle synergies and multimodal interfaces
Diego A. Galeano, F. Brunetti, D. Torricelli, S. Piazza, JLP Rovira
BioMed research international (Hindawi) 2014.
2013
A low cost platform based on FES and muscle synergies for postural control research and rehabilitation
Diego A. Galeano, F. Brunetti, S. Piazza, D. Torricelli
NEUROTECHNIX 2013.

1 - Predicting the frequencies of drug side effects

A central issue in drug risk-benefit assessment is the identification of the frequencies of side effects in humans. Currently, these frequencies are experimentally determined in randomised controlled clinical trials. We developed a novel machine learning framework for computationally predicting the frequencies of drug side effects. Our matrix decomposition algorithm learns latent signatures of drugs and side effects that are both reproducible and biologically interpretable. We show the usefulness of our approach on 759 structurally and therapeutically diverse drugs and 994 side effects from all human physiological systems. Our approach can be applied to any drug or compound for which a few side effect frequencies have been identified, in order to predict the frequencies of further, yet unidentified, side effects. We also found that our model is informative of the biology underlying drug activity: individual components of the drug signatures are related to the distinct anatomical categories of the drugs and to the specific drug routes of administration.

2 - Extending the druggable genome

The identification of unknown targets of drugs is critical for the development of novel medicines and for the repositioning of old drugs. Yet, the molecular characterisation of molecules remains a daunting task. Recent estimates indicates that the number of targets known for approved drugs is only 4.5 (DrugBank 5.1). The goal of this project is to develop a machine learning model that not only allow to predict novel drug targets but also, it can help us to extend the known druggable genome.
References
1.
Hopkins, A. L. & Groom, C. R. The druggable genome. Nature Reviews Drug Discovery 1, 727–730 (2002).
1.
Finan, C. et al. The druggable genome and support for target identification and validation in drug development. Science Translational Medicine 9, (2017).
1.
Russ, A. P. & Lampel, S. The druggable genome: an update. Drug Discovery Today 10, 1607–1610 (2005).
1.
Rask-Andersen, M., Masuram, S. & Schiöth, H. B. The Druggable Genome: Evaluation of Drug Targets in Clinical Trials Suggests Major Shifts in Molecular Class and Indication. Annu. Rev. Pharmacol. Toxicol. 54, 9–26 (2014).

Diego Galeano is a Postdoctoral Researcher in Data Science at Fundação Getulio Vargas (FGV) Rio de Janeiro, Brasil. Diego completed his Ph.D. at Royal Holloway, University of London in 2019, working under the supervision of Prof. Alberto Paccanaro. Diego received a full-ride ITAIPU scholarship for undergraduate studies in Paraguay, and in 2015, the BECAL scholarship for his doctoral studies in London. Diego was a research fellow of GersteinLab at Yale University in 2017, where he worked in model interpretability and enhancer prediction. This research visit was thanks to the Royal Holloway Travel Award, the Santander Travel Award and the BECAL Travel Award. He also received the best poster and presentation awards for three years at the Annual Computer Science Colloquium at Royal Holloway. His interests lie in applications of AI and machine learning in healthcare, biology and medicine. He is also interested in the social impact of AI.


FGVEmap

Posdoctoral Researcher
Escola de Matemática Aplicada, FGV-Emap
Office: NA

diego.galeano@fgv.br


Google Scholar


Many thanks to Will Hamilton for the site template