Project title: Revisiting the genetic replicability by a multi- and personomic approach to facilitate the design of personalized treatments based on the causes of the illness
PIs: Igor Zwir (delval@decsai.ugr.es) and Coral del Val (zwir@decsai.ugr.es) from the Mining for Modelling group at the University of Granada (Spain)
Application period: October 17th to November 7th of 2019
Keywords: biomedicine, data analysis, personalized medicine, molecular genetics
Requirements:
- Preferably Degree in Computer Science, or "Bio" Degrees with Master in Bioinformatics, Data Science or similar.
- Advanced knowledge in Python, R, or C. will be valued.
- Knowledge in bioinformatics, molecular biology and / or genomics will be valued
- Experience with clusters in GNU / Linux environments will be valued
Summary:
Identifying the risk of suffering a disease requires the understanding of the genetic architecture of such disease, as well as the specificity of the underlying phenotype. We previously hypothesized and showed that the missing heritability was not missing but distributed in partitions of a complex disease into collections of syndromes each with specific genetic associations. However, despite the multiples efforts towards bridging the gap between bench and bedside, genetic association studies (GWAS) of mental disorders have largely produced weak, inconsistent and/or unreplicable findings. Patients with mental disorders continue receiving the same diagnosis, yet sharing few symptoms in common that vary widely in severity. Currently used antipsychotic medications have serious side effects and only limited efficacy (largely limited to reducing delusions and hallucinations) or no efficacy (particularly over negative and/or cognitive symptoms). Therefore, the lack of replicability, and thus the uncertainty about the genetic causality or influence over the manifestation of a disease still hinders translational scientists and clinicians to design personalized treatments based on the causes of the illness.
In order to tackle this challenge, we now hypothesize that: “the genetic reproducibility (variability) is not missed but distributed into different levels of omic organizations. Indeed, we propose that the differential syndromes that decompose a disease and their genetic reproducibility may allow early predictions of the longitudinal evolution of the phenotype of a disease, as well as its late occurrence (e.g., dementias/Alzheimer)”. Accomplishing this proposal will benefit outcome prediction and person‑centered diagnosis and treatment. The aims here proposed will make possible the empirical identification of disease subsets on the basis of the intersection of multiple common omics that may lead to the development of new, more person-specific and thus more effective treatments. We expect that the new generated knowledge will provide better diagnoses, early detection and an immediate low-cost genetic test (estimated in one euro) that could be used in the praxis for accurate differentiation of the eight different Schizophrenias. We pretend to apply the new extracted knowledge to the development of a grapheme biochip that will facilitate the selection of the treatment or therapy to follow increasing the well –being of the patients.
Moreover, successful completion of this project will result in a completely new software platform to search for new query studies and understand them in terms of a data-driven evidence-based patients x features associations. This will allow understanding diagnostic boundaries in multi-domain datasets including structural imaging, structural and functional connectivity data, and behavioral/phenomenological/clinical data and the obtention of new knowledge. Such a set of tools should permit to increase the precision of diagnoses as well as the personalization of treatment of psychiatric disorders.