Pathway-based Cell Models and Text-Processing Techniques for Analysis and Simulation of Phenotypes with Application to Biomedicine

Date

May 9, 2023

Time

11:00 AM - 12:00 PM

Speaker

Alfredo Ferro and Salvatore Alaimo

Affiliation

Department of Clinical and Experimental Medicine, University of Catania, Italy

Language

en

Main Topic

Biologie

Host

Gene Myers

Description

This talk will summarize three main techniques we developed in our Bioinformatics Lab in Catania. 1) MITHrIL: This system uses the concept of a metapathway, a mechanistic model of the cell, merging KEGG and REACTOME extended with microRNA and TF in a multi-relational network. MITHrIL takes as input the Log-Fold-Changes of differentially expressed biological elements (genes, metabolites, miRNAs). It will produce a quantitative prediction of each gene and pathway perturbation with its statistical significance. 2) PHENSIM: This system uses the MITHrIL model to simulate phenotypes. Phensim takes as input user-specified positively and negatively perturbed elements (genes, metabolites, miRNAs) and non-expressed genes in the cellular context (i.e., tissue, cell line) and will produce a qualitative prediction of gene and pathway alteration (positive, negative, or null) with its statistical significance. 3) NETME: This algorithm builds a knowledge graph starting from bio-medical literature given as input by the user. Papers can be obtained directly by querying PubMed or as PDF documents. The knowledge graph built by NETME summarizes the knowledge contained in those documents. We will show examples of these techniques for drug suggestion in infection and other diseases for pandemic first-aid intervention and precision medicine. Other applications will be briefly discussed, such as gene knock-out and knock-down simulation and extra-cellular exosome functional prediction. Finally, possible future extensions to single-cell and cell-cell communications will be outlined.

Last modified: May 10, 2023, 7:38:57 AM