Related Experiment Video
Updated: Jan 13, 2026
09:28
Engineering Molecular Recognition with Bio-mimetic Polymers on Single Walled Carbon Nanotubes
Published on: January 10, 2017
8.5K
Machine Learning Prediction of Analyte-Induced Fluorescence Perturbations in DNA-Functionalized Carbon Nanotubes
Sayantani Chakraborty1, Andrew T Krasley2, Colby H Smith1
1Department of Chemistry and Biochemistry, The University of Texas at El Paso, El Paso, Texas 79968, United States.
Nano Letters
|January 7, 2026
Summary
Machine learning models predict how molecules affect DNA-SWCNT nanosensor fluorescence. This approach aids in designing better near-infrared sensors for molecular detection.
Area of Science:
- Nanotechnology
- Biomolecular Engineering
- Computational Chemistry
Background:
- Single-walled carbon nanotubes (SWCNTs) functionalized with DNA act as near-infrared nanosensors.
- Predicting analyte-specific optical responses for these nanosensors is difficult.
Purpose of the Study:
- To develop machine learning (ML) models for predicting analyte-induced fluorescence changes in DNA-SWCNT dopamine nanosensors.
- To identify structural features correlating with nanosensor optical response strength.
Main Methods:
- Utilized a dataset of 63 small molecules around dopamine.
- Encoded analytes using RDKit fingerprints, with and without HOMO/LUMO energies.
- Applied principal component analysis and trained support vector regression/classification models.
- Compared ensemble and cross-validation strategies for model training.
Main Results:
- Regression models yielded mean R² values of 0.2-0.4, with cross-validation superior to ensembles.
- Classification models achieved mean F1 scores of approximately 0.8.
- Cross-validation demonstrated the best predictive performance on an independent test set of 21 molecules.
Conclusions:
- Machine learning effectively captures structure-response relationships even with limited data.
- ML models can guide the in silico design of DNA-SWCNT nanosensors.
- This work advances the development of targeted molecular sensing technologies.
