#TiRNA: a coarse-grained method with temperature and ion effects for RNA structure folding and prediction

##Overview

TiRNA is a coarse-grained computational method developed by Tan-group at Wuhan University for RNA structure modeling. The model incorporates the effects of temperature and ions to provide predictions for 3D structures and thermal stability of RNAs in monovalent/divalent ion solutions.

###Key Features:

• (1) Three-bead coarse-grained representation (P, C4', and N1/N9);
• (2) Force field accounting for temperature and ion effects;
• (3) Replica-exchange Monte Carlo & Monte Carlo simulated annealing for structure sampling;

###Package Modules:

• (1) RNA 3D Structure Prediction: Predicting 3D structures from sequences or secondary structures at given ion conditions;
• (2) RNA Thermal Stability Prediction: Predicting structures at various temperatures which can be used for analyzing melting temperatures and thermally unfolding pathways.

##System Requirements

• Operating System: Linux;
• CPU: ≥8 cores with 2 threads;
• Compiler: GCC ≥7.5 (supporting C++11 standard);
• Python: Version ≥3.11.5;
• Python Packages: Biopython ≥1.80 (pip install biopython).

##Installation

1.Download the TiRNA package

• git clone https://github.com/Tan-group/TiRNA.git

2.Verify dependencies

• gcc --version # Should be ≥7.5;
• python --version # Should be ≥3.11.5;
• python -c "import Bio; print(f'Biopython version: {Bio.version}')" #≥1.70 .

Usage Examples

1.Prediction for RNA 3D Structure and Thermal Stability from Sequences

Location: Run in from-sequence/

Steps:

(1) Prepare sequence file (seq.dat)

• Input your RNA sequence in seq.dat
• Example:
• GGCGAUGUCCAGCAGAUACACGUCGUUCGCACC

(2) Configure parameters (config.dat)

• Sampling_type 1
• Folding_steps 750000
• Optimizing_steps 500000
• C_Na 1000
• C_Mg 0
• N_cout 10

(3) Run the simulation with TiRNA

• bash run.sh

2.RNA 3D Structure Prediction with Given Secondary Structure

• Location: Run in from-2D-structure/

Steps:

(1) Prepare sequence and secondary structure (seq.dat)

• Format: sequence followed by secondary structure in dot-bracket notation
• Example:
• GGAGGAAGGAGCCUCC
• (((((......)))))

(2) Configure parameters (config.dat) - same as those in from-sequence

(3) Run the simulation with TiRNA

• bash run.sh

3.RNA 3D Structure Prediction from Initial Structures

3A. From Initial Structure in PDB Format

• Location: Run in from-3D-structure/PDB
• Steps:
• (1) Place your initial PDB structure file in the directory;
• (2) Configure parameters in config.dat;
• (3) Run with bash run.sh.

3B. From 3D Structure in CIF Format

• Location: Run in from-3D-structure/CIF
• Steps:
• (1) Place your initial CIF structure file in the directory;
• (2) Configure parameters in config.dat;
• (3) Run with bash run.sh.

Configuration Parameters

Configuration Parameters

Parameter	Description
Sampling_type	1 for replica-exchange Monte Carlo (REMC),0 for Monte Carlo simulated annealing(MCSA)
Folding_steps	Number of steps for structure folding
Optimizing_steps	Number of steps for structure optimization
C_Na	Na⁺ concentration in mM
C_Mg	Mg²⁺ concentration in mM
N_cout	Number of output predicted 3D structures

Performance Recommendations:

For Replica-Exchange Monte Carlo (REMC, Sampling_type = 1):

• Folding_steps: ≥500,000 steps recommended;
• Optimizing_steps: ≥100,000 steps recommended.

For Monte Carlo Simulated Annealing (MCSA, Sampling_type = 0):

• Folding_steps: ≥2,000,000 steps recommended;
• Optimizing_steps: ≥100,000 steps recommended.

Application-specific recommendations

• For more accurate 3D structure prediction:
  • REMC: Use ≥750,000 folding steps and ≥500,000 optimizing steps;
  • MCSA: Use ≥3,000,000 folding steps and ≥500,000 optimizing steps.
• For more accurate thermal stability prediction:
  • REMC: Use ≥4,000,000 folding steps;
  • MCSA: Use ≥20,000,000 folding steps.

Output Files

After successful execution, results are saved in the results/ directory:

Directory	Content Description
Folding_trajectory/	Folding trajectories at different temperatures
CG_structures/	Predicted top-N 3D CG structures
Secondary_structure/	Predicted top-N secondary structures in dot-bracket notation
All_atom_structure/	All-atom structures corresponding to top-N CG structures
Thermal_Stability/	Data for thermal stability analysis

Software Usage Notes:

• Place all input files in the data directory before running the program.
• Only seq.dat is accepted as the input sequence file.
• The program will overwrite the result output folder on each run. To preserve previous results:
  • Rename the existing result folder, or
  • Move it to another location before execution.

Post-Processing

To refine the predicted all-atom structures and remove potential steric clashes or chain breaks, use QRNAS:

• Install QRNAS from: https://github.com/sunandan-mukherjee/QRNAS.git

Support and Contact

For questions or issues regarding TiRNA, please contact: zjtan@whu.edu.cn.

References

• [1] Wang X, Lou E, Yu S, Tan YL, Shi YZ, & Tan ZJ. 2025. TiRNA: a coarse-grained method with temperature and ion effects for RNA structure folding and prediction. In preparation.
• [2] Wang X, Tan YL, Yu S, Shi YZ, & Tan ZJ. 2023. Predicting 3D structures and stabilities for complex RNA pseudoknots in ion solutions. Biophys J. 122, 1503-1516.
• [3] Shi YZ, Wang FH, Wu YY, & Tan ZJ. 2014. A coarse-grained model with implicit salt for RNAs: Predicting 3D structure, stability and salt effect. J Chem Phys. 141, 105102.
• [4] Stasiewicz J, Mukherjee S, Nithin C, & Bujnicki JM. 2019. QRNAS: Software tool for refinement of nucleic acid structures. BMC Struct Biol. 19, 5.

TiRNA Package - Tan Group, Wuhan University