Logo image
Two applications of computation to problems in chemistry
Thesis   Open access

Two applications of computation to problems in chemistry

Tina Thanh Dinh
Master of Science (M.S.), Drexel University
Jun 2023
DOI:
https://doi.org/10.17918/00001685
pdf
Dinh_Tina_20233.47 MBDownloadView

Abstract

Chemical problems Computational Chemistry
In this work, two different problems in chemistry will be explored using two different applications of computation. The first problem involves a modification to a muscle-like acid base switchable [2](2)rotaxane reported by Li et al. [J. Org. Chem. 79(15): 6996-7004] for improved force delivery. Semi-empirical electronic structure theory calculations were used to explore the mechanical behavior and switchable nature of the system. The original system does not work as envisioned, therefore a modification to the system was proposed and similar semi-empirical calculations were carried out. This modified system could potentially serve as the fundamental element of a molecular muscle. The second problem explored is the possible use of vectors of toxicity descriptors to predict antiviral activity. Since toxicity is an indicator of bioactivity, it is hypothesized that vectors of numerical toxicity descriptors may be used as a predictor of antiviral activity. This can be tested by making a distribution of toxicity vectors for known antiviral compounds and comparing them to similar vectors for other toxic, non-toxic, and benign compounds as well as a small subset of known antivirals held out of the distribution defining set. The vectors can then be ranked based on how similar they are to those of the known antivirals. This method can be used as an efficient and inexpensive way to screen for antiviral activity by eliminating low probability compounds from a list of prospects, which could decrease the cost and duration of the drug discovery phase of antiviral development.

Metrics

54 File views/ downloads
74 Record Views

Details

Logo image