Klaus Bernhard Gerwert

This paper presents a computational method to calculate retinal isomerization in real time, revealing an asymmetric potential energy landscape and branched photocycle that align with experimental data.

The paper predicts high-accuracy vibrational spectra from molecular structures using simulations and evaluates methods for IR-based structure determination.

The paper validates a computational prediction of an antibody mutation that abolishes antigen binding using experimental methods, demonstrating its utility in Alzheimer's disease diagnostics.

The paper uses high-resolution imaging and biophysical methods to show that Lewy pathology in Parkinson's disease consists mainly of a crowded membranous medley of vesicles and dysmorphic organelles, with only a small fraction containing proteinaceous filaments.

This paper develops a deep learning method for preprocessing infrared spectral imaging data using stacked contractive autoencoders, enabling domain-adaptable classification across different tissue types.