Multiomics integration in PANS
Master Degree (Artificial Intelligence curriculum)
CHL course, Academic Year: 2021/2022
Date: May 2022
Description
Integrate metabolomics and proteomics data from PANS. Perform classification of patients based on the two types of data, and extract biomarkers that maximize classification performance.
Abstract
PANS is a pediatric disorder that involves symptoms such as obsessive-compulsive disorder (OCD), tics, anxiety attacks, depression, and sleep problems. The causes of the onset of this disease are probably related to an immune response to a bacterial or viral infection; therefore, current treatments include antibiotics, anti-inflammatories, and antidepressants. Our project is based on two data sets: metabolic data and proteomic data. Most of the patients they refer to are in common. Then, as we described in the Method section, we merged these two datasets into a new one to try the a-priori methodology. Hence, our aim is to integrate omics data to maximize the classification performance extracting the biomarkers. This report is organized as followi