Implementation of the RNA design methd proposed in the paper (ISMB 2023):
[1] Zhou, T., Dai, N., Li, S., Ward, M., Mathews, D.H. and Huang, L., 2023. RNA design via structure-aware multifrontier ensemble optimization. Bioinformatics, 39(Supplement_1), pp.i563-i571.
🚨 Important Notice: 🚨
The main branch is currently being upgraded. For the latest stable version of the code used in the paper published in ISMB2023, please refer to the ismb2023 branch.
python3
pip install requirements.txt
export VIENNAPATH=/path/to/ViennaRNA/lib/python3.9/site-packages
python main.py --online # plus other design options
Example:
$ echo "(((((......)))))" | python main.py --online --step 100
args:
Namespace(path='', object='pd', k=10, t=1, step=100, name='', init='cg', repeat=1, start=0, nomfe=False, nosm=False, bp=False, online=True)
(((((......)))))
(((((......)))))
seed_np: 2020
steps: 100, t: 1, k: 10, structured mutation: True, ensemble objective: position_ed_pd_mfe
name_pair: cg
pair_pool: ['CG', 'GC']
0 CGCGCAAAAAAGCGCG: 0.8093, 0.1907
1 CCGCGAAAAAACGCGG: 0.9173, 0.0827
2 CCCGGAAAAAACCGGG: 0.9410, 0.0590
3 GGCCGAAAAAACGGCC: 0.9451, 0.0549
4 CCCCCAAAAAAGGGGG: 0.9384, 0.0616
5 GCGGCAAAAAAGCCGC: 0.9739, 0.0261
6 GCCCCAAAAAAGGGGC: 0.9746, 0.0254
7 GCGCCAAAAAAGGCGC: 0.9739, 0.0261
8 GGGGCAAAAAAGCCCC: 0.9746, 0.0254
9 GCCCGAAAAAACGGGC: 0.9450, 0.0550
iter: 10 value: 0.0254 mfe count: 18 umfe count: 18 best iter: 0 improve: 0.00e+00
iter: 20 value: 0.0254 mfe count: 26 umfe count: 26 best iter: 0 improve: 0.00e+00
iter: 30 value: 0.0241 mfe count: 35 umfe count: 35 best iter: 28 improve: -1.28e-03
iter: 40 value: 0.0241 mfe count: 42 umfe count: 42 best iter: 28 improve: 0.00e+00
iter: 50 value: 0.0241 mfe count: 51 umfe count: 51 best iter: 28 improve: 0.00e+00
iter: 60 value: 0.0241 mfe count: 59 umfe count: 59 best iter: 28 improve: 0.00e+00
iter: 70 value: 0.0241 mfe count: 68 umfe count: 68 best iter: 28 improve: 0.00e+00
iter: 80 value: 0.0241 mfe count: 77 umfe count: 77 best iter: 28 improve: 0.00e+00
iter: 90 value: 0.0241 mfe count: 85 umfe count: 85 best iter: 28 improve: 0.00e+00
iter: 100 value: 0.0241 mfe count: 92 umfe count: 92 best iter: 28 improve: 0.00e+00
RNA sequence:
GCCCCGAAAAAGGGGC
ensemble objective: 0.02414993696568135
(((((......)))))
(((((......)))))
structure distance: 0
count of mfe solutsion: 92
count of umfe solutions: 92
[RNAStructure('GGGGCAAAAACGCCCC', 0.9745324428444898), RNAStructure('GGGGCAGAACAGCCCC', 0.9745537982728839), RNAStructure('GGGGCAAAGAAGCCCC', 0.9745522394520901), RNAStructure('GGGGCAGAAAAGCCCC', 0.9745541635522057), RNAStructure('GGCCCAAAAAAGGGCC', 0.974676524461917), RNAStructure('GCCCCAAUAAAGGGGC', 0.9745688330920909), RNAStructure('GCCCCAAAAAAGGGGC', 0.9745692982930593), RNAStructure('GCCCCGAAAAAGGGGC', 0.9758500630343186), RNAStructure('GGGGCAAAAAAGCCCC', 0.9745543077208556), RNAStructure('GGAGCGAAGAAGCUCC', 0.9756107700034071)]
mfe samples: ['GGGCUCAAGAUGGCCC', 'GCCCUUAAAACGGGGC', 'GCCCAAAAAACUGGGC', 'GGCACAGAAAAGUGCC', 'GCCCUAAUAACAGGGC', 'GGGUGCAAGACCACCC', 'GGGGCAAAGAAGCUUC', 'GGGGCAGAACAGCCCC', 'GGUGCAGAAAAGUGCC', 'GCCCCAAAAGAGGGGC']
umfe samples: ['GGGCUCAAGAUGGCCC', 'GCCCUUAAAACGGGGC', 'GCCCAAAAAACUGGGC', 'GGCACAGAAAAGUGCC', 'GCCCUAAUAACAGGGC', 'GGGUGCAAGACCACCC', 'GGGGCAAAGAAGCUUC', 'GGGGCAGAACAGCCCC', 'GGUGCAGAAAAGUGCC', 'GCCCCAAAAGAGGGGC']
kbest: [RNAStructure('GGGGCAAAAACGCCCC', 0.9745324428444898), RNAStructure('GGGGCAGAACAGCCCC', 0.9745537982728839), RNAStructure('GGGGCAAAGAAGCCCC', 0.9745522394520901), RNAStructure('GGGGCAGAAAAGCCCC', 0.9745541635522057), RNAStructure('GGCCCAAAAAAGGGCC', 0.974676524461917), RNAStructure('GCCCCAAUAAAGGGGC', 0.9745688330920909), RNAStructure('GCCCCAAAAAAGGGGC', 0.9745692982930593), RNAStructure('GCCCCGAAAAAGGGGC', 0.9758500630343186), RNAStructure('GGGGCAAAAAAGCCCC', 0.9745543077208556), RNAStructure('GGAGCGAAGAAGCUCC', 0.9756107700034071)]
full results are saved in the file: puzzle_[[[[[......]]]]]_seed_2020.json
python main.py --t 1 --k 10 --object pd --path data/eterna/eterna100.txt # plus other design options
--online turn on online mode.
--t set the sampling temprature(T in the paper). The default value is 1.
--k set the size of frontier(priority queue). The default value is 10.
--object set objective, the value can be pd or ned, which stand for probability defect and ensemble defect respectively. The default value is pd.
--init the sequence initilization method, and the value can be cg or all, which stand for targeted initilization and random initilization respectively. The default value is cg.
--nosm if used, the structured mutation will be ablated.
--nomfe if used, "mfe as product" will be ablated, i.e., only the sequence with the best score will be used as the designed sequence.
--step set maximum number of iterations. The default value is 5000.
--repeat set the number of repeated experiments. The default value is 1. (only for batch mode)
--start set the starting index for naming the output file. The default value is 0. (only for batch mode)
The design results will be output as a csv file with the following columns.
structure: the input puzzle (target secondary structure).
objective: the best (lowest) objective value achived during optimization.
rna: the sequence with the best objective during optimization.
mfe: the MFE structure of rna, given by ViennaRNA folding engine.
dist: the distance between mfe and structure.
mfe_list: a list containing all the MFE solutions found. The order of the list is based on the time when each MFE solution is detected.
umfe_list: a list containing all the uMFE solutions found. The order of the list is based on the time when each uMFE solution is detected.
k_best: the priority queue at the final iteration.
log: the objective values at each iteration.
time: the total time used to design the input puzzle structure.
see data/results/eterna_samfeo.csv
Blue denotes pairs in the target structure and Red denotes predicted pairs not in the target structure. The darkness of the curves indicates pairing probability. The following figs show the design results for the puzzle "Runner" by NEMO and SAMFEO respectively. We can see that there are some red curves in the plot of NEMO, which means those undesired pairs are more likely to apear in the ensemble of sequence designed by NEMO.
