FiBeFTa: Finding the best fingerprint for discerning between actives and decoys: application to the SARS-CoV-2 main protease

Identifier:  TFG:4468

Handle: https://hdl.handle.net/20.500.11797/TFG4468

Authors:  Villaró Serrano, Oriol

Abstract:
Molecular fingerprints have been used regularly in virtual screening and drug discovery. Molecular fingerprints combine the results of more complex techniques, paired with the efficiency that comes with binary data structures. However, there are different types with a huge variety within them, that each results in a different fingerprint for the same molecule. This project aims to develop a tool to compare 10 different fingerprints, and use it to rank how each of them performs according to 4 different metrics. In ordero to achieve this, we will use this tool on one of the common benchmarks that exist in cheminformatics, the Directory of Useful Decoys.