Michael Heinzinger

Michael Heinzinger is a researcher at the Chair for Bioinformatics within the School of Informatics at Technische Universität München. His work focuses on protein language models, machine learning applications in structural biology, and sequence-structure-function relationships. He contributes to projects like ProtTrans and participates in teaching activities including Data Mining and Problem-Based Learning (PBL) modules.

Research Focus: Heinzinger's research explores

• Protein language modeling and representation learning
• Structure prediction using deep learning
• Functional annotation through sequence embeddings
• Evolutionary insights via domain analysis
• Transmembrane protein visualization tools

Publication Trends: Recent work emphasizes protein language models (ProtTrans, HiFi-NN), structure prediction without alignments, binding residue analysis in disordered regions, and evolutionary studies of venom/toxin genes. Techniques include embeddings, contrastive learning, and attention mechanisms applied to protein space visualization and functional prediction.

Contact:

• Email: ga32bav@mytum.de