Chemická vzdálenost a její využití při vyhledávání molekul léků v databázích

Bc. Peter Kabát

Diplomová práce

Chemical distance and its application in searching of drug molecules in databases

Abstract:

This thesis is focused on computation of chemical distance between molecules. The reason of this computation is to find out, how similar/different molecules structures are.This can be helpful especially by the drugs, because the similar structures have usually very similar properties. Then I had to study all necessary chemical terms, which are explained in the text, from the basics to more advanced topics. I also have to get familiar with the SDF format, which was used in the data file, in order to completely understand the representation of molecules by text file. The implemented program, ChemDist, was created on the basis of the mathematical definition by Dugundji and Ugi, was very slow, what was causedby the extreme computational difficulty. That is why some optimizations were necessary in order to decrease the computational time and to make this application usable. In the end an approximative computational approach was proposed as the best solution. And this improvement was also implemented and tested on some sets of isomeric molecules. The results do not consist only of the values of chemical distances, but they also consist of molecular families, to which molecules were divided according to their value of chemical distance. TheresultsoftheapplicationChemDist canbefoundinthefollowingtext.

Keywords

Chemistry, Isomeria, Chemical distance, NP-complete problem, Java, SDF format, DTP NCI database, Molecular graph, Isomorphism, ChemDist

Jazyk práce: angličtina

Datum vytvoření / odevzdání či podání práce: 4. 1. 2007

Obhajoba závěrečné práce

• Obhajoba proběhla 14. 2. 2007
• Vedoucí: doc. RNDr. Radka Svobodová Vařeková, Ph.D.

Masarykova univerzita

Fakulta informatiky

Magisterský studijní program / obor:
Aplikovaná informatika / Aplikovaná informatika