Search

Browse

• Communities
• Authors
• Year
• Titles
• Subjects

My Space

• Login

Chemical similarity of cancer drugs and the construction of a comprehensive drug target network,focusing on the cancer metabolic pathway.

Kallergis Georgios

Full record

---

Summary

Drug Repositioning using in silico approaches, gains more and more groundboth in academia and in the R& D departments of big pharmaceutical companiesmostly because they result in a significant reduction to cost and timeassisting in designing more efficient and safer drug molecules. One of themain fields it has been deployed is treating and preventing cancer, whichdespite the massive efforts on time and investments has not been efficientlytackled yet. This has led to reposition compounds like Acetylsalicylic acid(Aspirin) and Metformin to cure various types of tumour. Chemical similarityis an important concept in drug research. In silico tools for similaritybasedapproaches are used to identify compounds with similar bioactivitiesbased on structural similarity between two ligands that could share sameor similar targets. As chemical similarity seems to be an indication for similarpharmacological activity, this thesis aims to take advantage of it anddevelop two computational approaches to find approved or experimentaldrugs which could possibly be used to treat cancer. These are based on algorithmscommon in chemoinformatics, on machine learning models, andensemble methods to enhance their performance in order to make confidentpredictions. The approaches are based on the assessment of the SimplifiedMolecular-Input Line-Entry System (SMILES) as adequate molecularstructure representations for the identification of drug similarities. The firstmethod aims to pairwise drug similarity or similarity among a few compounds,whereas the second one focuses on drug group similarity. Resultsinclude suggestions of coxib similar drugs and repositioned drugs focusingon metabolic pathways of cancer.

Available Files

Services