Panagiotis Antoniadis

Decoding biological sequences is crucial as it reveals the biological functions associated with these sequences, enabling scientists to uncover the genetic basis of diseases, identify drug targets and predict protein function. However, analyzing biological data is extremely challenging due to their high complexity and variability; even minor changes in the sequence can significantly impact biological functions. Also, biological data are often noisy and lack proper annotation. The goal of this PhD project is to develop deep learning methods that can capture the complex relationships in biological sequences. Towards this end, we are going to implement foundational models trained on multi-modal biological data using self-supervised learning techniques. We will also investigate the integration of data from simulations and experimental results into our proposed learning methods.