Research Project

Project Title:

USE OF ARTIFICIAL INTELLIGENCE AND SYSTEMS BIOLOGY FOR DIAGNOSIS AND PERSONALIZED RISK ASSESSMENT FOR INHERITED KIDNEY DISEASES, FOCUSING ON ALPORT SYNDROME

Project Type:

Translational research

Disease group(s):

Hereditary glomerulopathies, Tubulopathies, Metabolic & stone disorders, Thrombotic microangiopathies, AD structural kidney disorders

Project Summary:

BACKGROUND:
Inherited kidney diseases (IKD) are the leading cause of dialysis or kidney transplant in childhood, and in 10% of adults who need these techniques to survive. IKD diagnosis is very challenging, since there are more than 300 IKD and they are largely unknown to the nephrologist. Alport Syndrome is the second most common IKD and an excellent model of IKD that causes kidney failure, hearing loss and vision problems.
Machine learning approaches in nephrology are not yet as explored as in other medical fields due to the complexity of kidneys. The discipline of network medicine, that brings together systems biology and network science, provides a solid biological framework in which omics data and other clinical data can be integrated to provide novel diagnosis and risk stratification strategies. None of these techniques has been applied to IKD so far.
MAIN OBJECTIVES:
Design a machine learning tool that facilitates the diagnosis of IKD
Develop a prediction tool that integrates clinical and Omics data for Autosomal dominant Alport syndrome (ADAS).
METHODOLOGY:
Cohort of 280 fully phenotyped patients with ADAS
Machine learning for diagnosis of IKD: Structured representation of kidney disease and development and deployment of computer-assisted diagnosis algorithms. Development of a web-based tool for automatic diagnosis and contribution to the community
Genomics: Sequencing of capture-based 316 IKD gene panel
Urine analysis: Non-biased assessment of urine peptidomics (capillary electrophoresis–mass spectrometry) and metabolomics (NMR and validation by Selected Reaction Monitoring liquid chromatography-triple quadrupole mass spectrometry). Hypothesis-driven: ELISA assays for Klotho and DKK-3
Epigenetics: Analysis of both the genotype and DNA methylation in regions of interest using the Genome and Epigenome Unified Sequencing (GEUS).
Advanced statistical analysis. Identification of risk patterns using machine learning methods
EXPECTED RESULTS
This project aims to facilitate the diagnosis of IKD, by means of an on-line tool and, using the model of Autosomal Dominant Alport Syndrome, to obtain a personalized prediction tool.

Lead principal investigator(s):

Roser Torra, Barcelona

Project Period:

06/2021   -   06/2025

Sponsors:

National funding agency

« Back to research page