Speaker
Description
Self-assembling peptides (SAPs) are a type of biomaterial consisting of short aminoacid sequences that can be controlled under specific physicochemical conditions. SAPs form nanostructures that can mimic biological scaffolds giving them numerous applications such as in drug delivery, tissue engineering, biosensors, etc.
In this project we will create new SAP sequences based on desired biophysicochemical properties. Not only has deep learning proven adept at this task, it may also help us shed some light on the correlation between peptide sequence, structure and biological activity.
However the quality of the result heavily depends on the number and quality of the training data. It is therefore paramount to construct a database of known peptide sequences and their biophysicochemical properties obtained by our and other groups from wet lab experiments and complemented with computational methods. The database should be open and possess a user-friendly interface adapted to professionals in the fields of chemistry and biology that might not have experience in working with extensive datasets. It should deal with both entries from automated experimental setups, as well as manual input from a variety of users. The data structure should be consistent. Changes to specific data may only be carried by the user who originally uploaded the data and people whom the user specified as collaborators. These data changes must be traceable.