The identification of robust lists of molecular biomarkers related to a disease is a fundamental step for early diagnosis and treatment. However, methodologies for biomarker discovery using high-throughput genomic data often provide results with limited overlap.
It has been suggested that one reason for these inconsistencies may be that in complex diseases, multiple genes belonging to one or more physiological pathways are associated with the outcome. Thus, a possible approach to improve list stability is to integrate biological information from genomic databases in the learning process.
In this talk, I will discuss the development and application of advanced machine learning and modeling methods for high-throughput biological data analysis in the field of Bioinformatics and Systems Biology. In particular, I will discuss the reasons of biomarker list instability and present: i) a method to regularize the discriminant functions of the classification algorithms using prior biological knowledge; ii) a method to identify single nucleotide polymorphisms (SNPs) significantly associated with a phenotype within predefined sets of SNPs such as pathways or genomic regions.
Finally, I will show how the information on the identified biomarkers can be integrated at different levels with transcriptomic and protein signaling data, to better understand the mechanisms beyond the pathogenesis of a disease.Contact person: Nicola Bombieri
CSS e script comuni siti DOL - frase 9957