Doctoral Dissertation: M.Sc. Fatemeh Seyednasrollah, February 4th, 2022

    Doctoral Dissertation: M.Sc. Fatemeh Seyednasrollah, February 4th, 2022

    MSc. Fatemeh Seyednasrollah presents her thesis “Machine Learning and Computational Methods to Identify Molecular and Clinical Markers for Complex Diseases – Case Studies in Cancer and Obesity” for public examination in the University of Turku on Friday February 4th at 12:00. Audience can follow the defence remotely: https://utu.zoom.us/j/68995818293

    Professor Ewa Szczurek (University of Warsaw) has been nominated as the opponent, and Professor Laura Elo (University of Turku) as the custos. The field of study is computational biology and medical biology and the language of the dissertation will be English.

    In her doctoral dissertation, Fatemeh Seyednasrollah developed mathematical tools to improve treatment decision making in complex diseases with case studies in cancer and obesity. The proposed predictive methods are utilized to estimate the disease onset, prognosis and to adopt beneficial therapy at individual patient level. Most often, clinical decision making relies on assessment of the disease sub-type and prognosis which can be a critical task without robust predictive methods. These predictive methods determine “which and how” a set of pre-defined bio-clinical markers can be utilized to assist in choosing the most appropriate therapeutic strategies.

    In case of renal cell carcinoma (the most common type of kidney cancer), tumors are proved to be heterogenous with distinct histological sub-types. This dissertation proposes a data driven risk stratifying method to identify prognostic biological markers for clear cell renal cell carcinoma using patients’ gene expression profiles. In addition to making a prognosis assessment, some of the proposed novel biomarkers have the potential to serve as drug target candidates.

    This dissertation also addresses the challenging question of predicting the efficacy of chemotherapy in metastatic prostate cancer. In clinical practice, not every patient with advanced prostate cancer would benefit from chemotherapy and instead, 10-20 percent of patients would develop undesirable adverse events and may face deterioration of survival time and life quality. As a solution, we proposed an ensemble based predictive model to assist clinicians to choose correct candidates for further chemotherapy treatments utilizing baseline clinical characteristics, Seyednasrollah explains.

    The thesis has been published online: https://www.utupub.fi/handle/10024/153083


    Feb 04, 2022 12:00