This is the third and last webinar of the series about the impact of AlphaFold on training and research in life sciences. This webinar will focus on the impact of AlphaFold on research and development in life sciences by highlighting a variety of case studies from academia as well as industries.

In this webinar we had the following panel of speakers:

Jan Kosinski - AlphaFold has revolutionised structural biology by enabling the modelling of monomeric proteins and small complexes at accuracy similar to experimental structures. However, to model large complexes, it still needs to be integrated with experimental data such as cryo-electron microscopy maps. I presented how such an integrative combination enabled us to model the human nuclear pore complex at unprecedented completeness and precision. I used this example to highlight the new opportunities for integrative structural biology. 

Juan Carlos Mobarec - Methods and software to obtain information about protein structures from amino-acid sequences have been applied in the pharmaceutical industry over the span of decades. Hence, what can deliver the AI-containing modelling workflows such as AlphaFold to early stage R&D? We have pioneered the application of AlphaFold in R&D on multiple cases. Here, we reviewed use case examples of the latest automated workflow for protein structure prediction.

Juri Rappsilber - Crosslinking mass spectrometry is now a well-established experimental technique for investigating protein interactions, structure and function. Crosslinking MS also remains unparalleled in its ability to provide structural information in complex systems. This is particularly relevant in the light of the most recent revolutionary advancement in structural biology. Artificial intelligence program-based predictions of protein structure (e.g. by AlphaFold) are having a wide-ranging impact on our ability to model protein structures but also to study protein interactions that are yet to be fully understood. Crosslinking mass spectrometry offers here highly relevant experimental data. Using a crosslinking mass spectrometry study of Bacillus subtilis we have investigated different aspects of joining AlphaFold and crosslinking. Particularly, the validity of the ipTM score was probed by crosslinking mass spectrometry, and we present the possibility of assembling larger complexes based on the pairwise data resulting from crosslinking mass spectrometry and the AlphaFold prediction of two-protein interactions. Ultimately, our structural understanding of biology stands to gain much from combining AlphaFold and crosslinking data, particularly in our drive towards building a structural and functional understanding of the cell.

Katja Luck - AlphaFold’s ability to predict the structure of folded regions of proteins and to identify disordered regions of proteins with unprecedented accuracy transforms experimental design in molecular biology. First, I discussed how AlphaFold can advance the design of functional and well-expressed protein constructs for the study of protein interactions and protein binding regions. Second, I highlighted the potential, possible approaches but also limitations of AlphaFold for the identification of protein-protein interaction interfaces.

Panel discussion was chaired by Janet Thornton.