Statistical study of chemical structure similarity based on binary data
International Journal of Computing and Optimization · 2015
DOI: 10.12988/ijco.2015.5826 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Computing and Optimization, 2015. Verfasst von Rizal Bakri und 2 weiteren Autor:innen.
Abstract (Originalsprache)
Similarity measurement of the chemical structure of a compound refers to the similarity of chemical elements, molecules or chemical compounds. It is assumed that similar compounds tend to have similar properties. From statistical point of view, chemical structure similarity measurement is essentially a measurement of object similarity based on binary variable because the chemical structure of a compound is represented in the molecular fingerprint. Tanimoto coefficient is most often used in the measurement of similarity among chemical structure of compounds. However, there are many other coefficients that can be used as alternative to the Tanimoto coefficient. In this study, we have collected 79 binary similarity coefficients from various literature. The result showed that Forbes-2 coefficient gives the higher performance than Tanimoto coefficient based on AUC value and the Simpson, the Yulew, and the Var of Correlation coefficients are also an alternative measure based on hierarcical cluster analysis.
Zitieren
APA
Bakri, R., Wijayanto, H., & Afendi, F. M. (2015). Statistical study of chemical structure similarity based on binary data. International Journal of Computing and Optimization. https://doi.org/10.12988/ijco.2015.5826
Vancouver
Bakri R, Wijayanto H, Afendi FM. Statistical study of chemical structure similarity based on binary data. International Journal of Computing and Optimization. 2015. doi:10.12988/ijco.2015.5826
BibTeX
@article{Bakri2015,
author = {Bakri, Rizal and Wijayanto, Hari and Afendi, Farit Mochamad},
title = {Statistical study of chemical structure similarity based on binary data},
journal = {International Journal of Computing and Optimization},
year = {2015},
doi = {10.12988/ijco.2015.5826}
}
EUVIMED