• A
  • A
  • A
  • ABC
  • ABC
  • ABC
  • А
  • А
  • А
  • А
  • А
Regular version of the site

Program for Analysis of 3D Molecular Fragments for Prediction of Properties

Student: Solovev Andrei

Supervisor: Rimma Akhmetsafina

Faculty: Faculty of Computer Science

Educational Programme: Software Engineering (Bachelor)

Year of Graduation: 2016

This paper deals with the development of new 3D fragment descriptors for building Quantitative Structure Property Relationship (QSPR) models and creation on their basis of the software product for modeling and prediction of properties. Known 2D fragment descriptors are one of the most important types of descriptors used for design of substances with the desired properties by QSPR modeling. An important disadvantage of 2D fragment descriptors is problems with the description of stereoisomeric molecules and conformers. In this paper, it is overcome by developing 3D fragment descriptors taking into account spatial arrangement of fragment atoms. The paper specifies universality of new fragments descriptors for model developments by different machine learning methods and usefulness of the program for researchers and in assessment of compound's properties. Functional purpose of developments is evaluation and prediction of physical, chemical properties and biological activities of substances including stereoisomers conformers based on experimental data about know organic molecules and their properties. In this work, new 3D fragment descriptors and algorithms for their generation and coding are developed. Forward and backward variable selection algorithms are used for selection of the most significant descriptors for QSPR models on the base of new descriptors. The known simplex differential evolution algorithm was adapted for selection of the most relevant fragment descriptors. Multiple Linear Regression as a machine learning method using of reliable and efficient singular value decomposition was implemented. The program for QSPR modeling with graphical interface and 3D representation of molecules was elaborated. An evaluation the effectiveness of new descriptors was performed by properties predictions of known compounds. Selected software tools, operational purpose of the program, its field of application and its limits are described.

Student Theses at HSE must be completed in accordance with the University Rules and regulations specified by each educational programme.

Summaries of all theses must be published and made freely available on the HSE website.

The full text of a thesis can be published in open access on the HSE website only if the authoring student (copyright holder) agrees, or, if the thesis was written by a team of students, if all the co-authors (copyright holders) agree. After a thesis is published on the HSE website, it obtains the status of an online publication.

Student theses are objects of copyright and their use is subject to limitations in accordance with the Russian Federation’s law on intellectual property.

In the event that a thesis is quoted or otherwise used, reference to the author’s name and the source of quotation is required.

Search all student theses