Researchers and users of computer programs that are currently applied to find amyloid structures in the Protein Data Bank (PDB) entries mostly have to manage at least one of the following disadvantages:
- The program is aimed to find only complete amyloid structures, therefore many entries can be ignored even if the majority of their structures resemble amyloids.
- The program uses connotation and other non-quantitative text data of PDB which in many cases cannot be considered as well-authenticated.
Our new program is constructed to avoid these problems mentioned above: using data mining methods, it can identify amyloid-like segments of PDB entries which are not necessarily complete amyloid structures, only from the available geometric data of an entry. In our opinion, our program can help in finding PDB entries which can be related to some neurodegenerative diseases (e.g., Alzheimer's disease).
In our poster, work versions of our program are presented as well as the structure of the final version. Figures and tables are used to visualize the developing steps and motivations behind the applied methods.