TAPERSS-MoCa: A Physics-Based Monte Carlo Framework for 3D RNA Structure Prediction from Sequence
4 affiliations
- 1Department of Chemistry and Biochemistry
- 2Florida Atlantic University
- 3Department of Chemistry
- 4Massachusetts Institute of Technology
Abstract
Ribonucleic acid (RNA) function is intrinsically linked to its three-dimensional architecture, yet experimental structure determination remains challenging. Here, we introduce TAPERSS-MoCa (Theoretical Analyses, Prediction, and Evaluation of RNA Structures from Sequence using a Monte-Carlo Simulator), a physics-based fragment assembly framework for predicting RNA tertiary structures directly from sequence. By utilizing a novel fragment library grounded in dinucleoside monophosphate (DNMP) energetics parameterized via Discrete Path Sampling (DPS), TAPERSS-MoCa efficiently samples low-energy conformational landscapes. Evaluated across 6-, 10-, and 12-nucleotide RNA test sets, TAPERSS-MoCa consistently recovers native-like global minima with canonical Watson–Crick base pairs adopting near-planar geometry. By combining dinucleoside fragment energetics with an angularly constrained hydrogen-bonding potential, TAPERSS-MoCa accurately captures A-form structures in 6-nucleotide linear RNAs, outperforming traditional fragment-assembly algorithms that frequently misfold these flexible sequences. Thermodynamic evaluations based on ensemble partition functions (Δ G folding ) and macrostate basin free energies (Δ F folding ) demonstrate strong predictive performance for rigid, well-stacked GNRA tetraloops. Furthermore, TAPERSS-MoCa systematically reproduces key physical stability trends predicted by empirical nearest-neighbor models. This physics-based approach provides a robust platform for structural prediction and thermodynamic profiling of short RNA motifs, supporting RNA-targeted drug discovery and functional design.
Related
- KAT-Net: a hyperbolic routing framework for joint lysine acylation site prediction with shared sequence–structure representations
- GeoEPred: A multimodal structure-aware geometric deep learning framework for Gram-negative bacterial secreted effector prediction with sequence semantics
- Representation learning for low-data protein fitness prediction using sequence–structure embeddings and graph neural networks
- GeoDSSOP: Geometry-aware prediction of protein backbone order parameters from sequence and structure
- AIPFusion-Net: an interaction-fusion framework for anti-inflammatory peptide prediction using protein language model embeddings, local sequence patterns, and handcrafted descriptors.
BibTeX
@misc{ibrahim2026taperssmoca,
title = {{TAPERSS-MoCa: A Physics-Based Monte Carlo Framework for 3D RNA Structure Prediction from Sequence}},
author = {Mariam Ibrahim and Ivan Issac Riveros and Nanayakkara P.N.R. Kumarachchi and Ilyas Yildirim},
howpublished = {ChemRxiv},
year = {2026},
month = oct,
doi = {10.26434/chemrxiv.15010570/v1},
url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15010570/v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - TAPERSS-MoCa: A Physics-Based Monte Carlo Framework for 3D RNA Structure Prediction from Sequence
AU - Mariam Ibrahim
AU - Ivan Issac Riveros
AU - Nanayakkara P.N.R. Kumarachchi
AU - Ilyas Yildirim
PY - 2026
DA - 2026/10/16
PB - ChemRxiv
DO - 10.26434/chemrxiv.15010570/v1
UR - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15010570/v1
AB - Ribonucleic acid (RNA) function is intrinsically linked to its three-dimensional architecture, yet experimental structure determination remains challenging. Here, we introduce TAPERSS-MoCa (Theoretical Analyses, Prediction, and Evaluation of RNA Structures from Sequence using a Monte-Carlo Simulator), a physics-based fragment assembly framework for predicting RNA tertiary structures directly from sequence. By utilizing a novel fragment library grounded in dinucleoside monophosphate (DNMP) energetics parameterized via Discrete Path Sampling (DPS), TAPERSS-MoCa efficiently samples low-energy conformational landscapes. Evaluated across 6-, 10-, and 12-nucleotide RNA test sets, TAPERSS-MoCa consistently recovers native-like global minima with canonical Watson–Crick base pairs adopting near-planar geometry. By combining dinucleoside fragment energetics with an angularly constrained hydrogen-bonding potential, TAPERSS-MoCa accurately captures A-form structures in 6-nucleotide linear RNAs, outperforming traditional fragment-assembly algorithms that frequently misfold these flexible sequences. Thermodynamic evaluations based on ensemble partition functions (Δ G folding ) and macrostate basin free energies (Δ F folding ) demonstrate strong predictive performance for rigid, well-stacked GNRA tetraloops. Furthermore, TAPERSS-MoCa systematically reproduces key physical stability trends predicted by empirical nearest-neighbor models. This physics-based approach provides a robust platform for structural prediction and thermodynamic profiling of short RNA motifs, supporting RNA-targeted drug discovery and functional design.
ER -