After graduating from Sakarya University Faculty of Medicine in 2020, I continued my academic work at Karadeniz Technical University, where I am currently a resident in the Department of Medical Pharmacology. My research interests took shape between the bedside applications of clinical pharmacology and the methods of computational drug discovery; I try to pursue these two areas as lines of work that feed into each other.

A substantial part of my clinical work takes place in the teratogenicity outpatient clinic, where I carry out risk assessment and counselling for patients presenting after exposure to drugs, chemicals or radiation during pregnancy and breastfeeding. Because the evidence base in this field is often limited, heterogeneous and drawn from data that are not specific to pregnancy, clinical decisions require the available literature to be interpreted case by case. For this reason, compiling exposure data systematically and making it more accessible to clinicians is also among the topics I work on.

My experimental line of research focuses on the behavioural consequences of metabolic disease. My core questions concern the effects of obesity, insulin resistance and related metabolic disorders on anxiety-like behaviour, depressive behaviour and learning and memory performance, and whether these effects can be modulated pharmacologically. Within this scope I work on study designs in which behavioural tests are evaluated together with biochemical and molecular measurements; my aim is to address the relationship between metabolic and neuropsychiatric findings across several measurement levels rather than at a single one.

My third area of work is AI-assisted drug research. Here I am building a discovery pipeline that runs from disease and target selection to de novo molecule generation, and from ADMET and central nervous system permeability filters to structure-based binding prediction and the novelty assessment of candidate molecules. The work is ongoing; so far I have focused mainly on defining the steps of the pipeline, keeping intermediate outputs auditable, and making the method applicable to different protein targets. My next goals include testing the generated candidates experimentally and evaluating the approach on a broader set of targets.

In parallel, I am interested in AI applications that support the research process itself. I try to build workflows for literature search, data analysis and scientific writing that preserve source accuracy and allow outputs to be audited retrospectively. I regard these tools not as systems that replace the researcher, but as an assisting layer that increases the speed and scope of work without compromising verifiability. In the longer term, my aim is to establish a way of working that turns questions arising from clinical practice into hypotheses testable by computational methods.