A General Computational Framework for Sequence-Level Thermodynamic Analysis of Modeled RNA-DNA and DNA-DNA Duplex States: A Proof-of-Concept Study in HIV-1
- 1Institute for Research in Fundamental Sciences
Abstract
RNA-DNA hybridization is relevant to several molecular processes, but computational energy estimates alone cannot establish hybrid formation in cells. This study presents a computational framework for comparing model-derived RNA-DNA and DNA-DNA energy metrics across 16 selected regions of the HIV-1 reference genome HXB2 (GenBank accession K03455.1). The operational metric was defined as the difference between the RNA-DNA calculation and the DNA-DNA reference calculation (ΔΔG_model). Across the 16 regions, 15 yielded negative metric values and one yielded a positive value. The mean was −8.00 kcal/mol (sample standard deviation, 5.45 kcal/mol). A two-sided one-sample t-test against zero gave t(15) = −5.87, p = 3.08 × 10⁻⁵. The mean remained negative in sensitivity analyses excluding either or both of the overlapping TAR and R5LTR regions. These findings describe a consistent direction in the selected computational panel. However, the regions are drawn from a single reference genome, are not independent biological replicates, and may not represent independent sequence information. Furthermore, our independent verification confirmed that while the RNA-DNA energies are highly reproducible, the DNA-DNA reference energy is highly dependent on the specific algorithmic approach. Eleven of the 16 DNA-DNA reference values showed numerical agreement with standard DuplexFold calculations, while five regions showed discrepancies (1.1–5.3 kcal/mol) that remain unresolved. The results do not demonstrate intracellular RNA-DNA hybrid formation or genomic R-loop formation. Instead, they provide a testable computational hypothesis for further evaluation through reproducible energy calculations and independent experimental measurements.
Related
- Thermodynamic regularization based computational framework for strain-induced crystallization
- EquiDesign: a versatile computational framework for de novo protein sequence design.
- The Self-and-Self Model: A General Computational Framework for Adaptive Selection and Coordination Among Internal Computational Mechanisms
- The Self-and-Self-and-World Model: A General Computational Framework for Adaptive Coordination Between Internal Computational Mechanisms and World-State Modeling
- The Self-and-Mutual-and-World Model: A General Computational Framework for Adaptive Coordination Between Autonomous Computation, Interactive Mechanisms, and World-State Modeling
BibTeX
@misc{samadi2026general,
title = {{A General Computational Framework for Sequence-Level Thermodynamic Analysis of Modeled RNA-DNA and DNA-DNA Duplex States: A Proof-of-Concept Study in HIV-1}},
author = {Saadat Samadi},
howpublished = {Zenodo},
year = {2026},
month = oct,
doi = {10.5281/zenodo.23615302},
url = {https://zenodo.org/records/23615302},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - A General Computational Framework for Sequence-Level Thermodynamic Analysis of Modeled RNA-DNA and DNA-DNA Duplex States: A Proof-of-Concept Study in HIV-1
AU - Saadat Samadi
PY - 2026
DA - 2026/10/24
PB - Zenodo
DO - 10.5281/zenodo.23615302
UR - https://zenodo.org/records/23615302
AB - RNA-DNA hybridization is relevant to several molecular processes, but computational energy estimates alone cannot establish hybrid formation in cells. This study presents a computational framework for comparing model-derived RNA-DNA and DNA-DNA energy metrics across 16 selected regions of the HIV-1 reference genome HXB2 (GenBank accession K03455.1). The operational metric was defined as the difference between the RNA-DNA calculation and the DNA-DNA reference calculation (ΔΔG_model). Across the 16 regions, 15 yielded negative metric values and one yielded a positive value. The mean was −8.00 kcal/mol (sample standard deviation, 5.45 kcal/mol). A two-sided one-sample t-test against zero gave t(15) = −5.87, p = 3.08 × 10⁻⁵. The mean remained negative in sensitivity analyses excluding either or both of the overlapping TAR and R5LTR regions. These findings describe a consistent direction in the selected computational panel. However, the regions are drawn from a single reference genome, are not independent biological replicates, and may not represent independent sequence information. Furthermore, our independent verification confirmed that while the RNA-DNA energies are highly reproducible, the DNA-DNA reference energy is highly dependent on the specific algorithmic approach. Eleven of the 16 DNA-DNA reference values showed numerical agreement with standard DuplexFold calculations, while five regions showed discrepancies (1.1–5.3 kcal/mol) that remain unresolved. The results do not demonstrate intracellular RNA-DNA hybrid formation or genomic R-loop formation. Instead, they provide a testable computational hypothesis for further evaluation through reproducible energy calculations and independent experimental measurements.
ER -