Using Explainable Artificial Intelligence and Molecular Dynamics to Investigate Sequence Dependent Fluorescence in Single Stranded DNA
Year: 2026
Type: preprint
Abstract: The intensity of a photon released from cyanine dye on a labeled oligonucleotide is heavily influenced by the adjacent sequence. This phenomenon could have profound effects in biotechnologies that rely on fluorescently labeled DNA based materials such as PCR, DNA-PAINT, and FISH. For instance, the selection of a DNA probe could result in a lower Signal-to-Noise ratio necessitating further data processing procedures. Currently, π- π stacking between the cyanine dye and the terminal nucleotide is the leading theory to explain this sequence dependent fluorescence. This stabilization minimizes the rotational isomerization of the cyanine dye, increasing the rate of photonic emissions. However, how this occurs in single stranded DNA with minimal secondary structure formation is not well understood. In this work, we use eXplainable Artifical Intelligence (XAI) and Molecular Dynamics (MD) to explore the potential mechanism of the phenomenon in addition to investigating some fundamental questions in biological data science.
Source: ChemRxiv
Authors: Micheal B. Reed, Reza Zadegan
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7165181667