Designing nano-systems for anticancer purposes by applying Perturbation Theory Machine Learning (PTML) models

Ricardo Santana; Robin Zuluaga; Piedad Gañán; Sonia Arrasate; Enrique Onieva; Humbert González-Díaz

Previous Article in event

PTML-LDA model applied to allosteric modulators

Previous Article in congress

Modeling and simulation on hydrogen-rich syngas production via gasification of palm kernel shell

Next Article in event

Alternative therapies for Mexican Leishmania.

Next Article in congress

PTML model of CHEMBL neurological diseases assays vs. protein sequence, and protein interaction networks in different brain regions

Designing nano-systems for anticancer purposes by applying Perturbation Theory Machine Learning (PTML) models

^{*

1},

²,

²,

³,

Enrique Onieva

⁴,

Humbert González-Díaz

⁵

¹ University of Deusto / Universidad Pontificia Bolivariana
² Universidad Pontificia Bolivariana
³ University of Basque Country
⁴ University of Deusto
⁵ Department of Organic Chemistry II and Basque Center for Biophysics (CSIC-UPVEHU), University of Basque Country (UPVEHU), Basque Country, Spain

Published: 13 June 2020 by MDPI in MOL2NET'20, Conference on Molecular, Biomed., Comput. & Network Science and Engineering, 6th ed. congress USEDAT-06: USA-Europe Data Analysis Training Program Workshop, Bilbao, Spain-Cambridge, UK-Miami, USA, 2020

https://doi.org/10.3390/mol2net-06-06863 (registering DOI)

Abstract:

The number of possible designs of nano-systems is elevated. The design depends on the function we need to develop. Among these systems we highlight Nanoparticle Drug Delivery Systems (DDNS) of high interest not only for Nanotechnology but also for Biomaterials science.^1–3

In this work we fusion the following information: 1) Drug-vitamin release nano-systems (DVRNs). This data set was collected from literature. 2) Vitamin derivatives data set extracted from ChEMBL database. Both data sets contain different assay conditions and molecular descriptors. Once we fusion the information, we apply Perturbation Theory Machine Learning (PTML) method in order to build the model. Once built with Perturbation Theory Operators (PT Operators), it presents both Specificity and Sensibility higher than 80%.

Until the best of our knowledge, we developed the first multi-label PTML model useful to design DVRNs for optimal biological activity.

Keywords: Nanotechnology; DVRN; Drug Delivery; Machine Learning; PTML

View paper

115 Reads
0 Recommendations

Comments on this paper

Viviana Quevedo

12 January 2021

Commentary

The methodology for developing the first multi-label PTML model to desing DVRNs is excellent. The novel research work as part of the literature review appears very robust