Addition/Correction Cite This: J. Phys. Chem. Lett. 2018, 9, 1674−1675
pubs.acs.org/JPCL
Correction to “Free Energy Landscape of GAGA and UUCG RNA Tetraloops” Sandro Bottaro,* Pavel Banás,̌ Jiří Šponer, and Giovanni Bussi* J. Phys. Chem. Lett. 2016, 7 (20), 4032−4038; DOI: 10.1021/acs.jpclett.6b01905
T
he results presented in our original paper1 were mistakenly computed using the histograms over the whole trajectory reweighted using the bias from metadynamics accumulated during the first 200 ns of each simulation only (i.e., one-fifth of the simulation length). We have recalculated the free energy differences using the bias from the entire trajectories (1 μs), obtaining different results. In particular, the new folding free energy for the ccGAGAgg tetraloop is 16.1 ± 0.5 kJ/mol and 22.8 ± 0.5 kJ/mol for ccUUCGgg. The difference with respect to the original paper is very small for the 6-mers (ΔΔG ≈ 1 kBT) and more significant for the 8-mers (ΔΔG ≈ 3 kBT), in agreement with the fact that longer oligomers require a longer time in order to report a statistically robust result. Despite this difference, all the conclusions drawn in the original paper remain valid. For completeness, we report here the corrected Figures 1−5. We are grateful to Andrea Cesari for identifying the error.
Figure 2. Free energy surfaces projected onto the eRMSD from native and onto the end-to-end distance at 300.9 K for UUCG tetraloop. In the two-dimensional projection, the colors indicate the free energy difference with respect to the minimum. Labels of the isolines are expressed in kBT. Gray shades indicate statistical error.
Figure 1. Free energy surfaces projected onto the eRMSD from native and onto the end-to-end distance at 300.9 K for GAGA tetraloop. In the two-dimensional projection, the colors indicate the free energy difference with respect to the minimum. Labels of the isolines are expressed in kBT. Gray shades indicate statistical error.
■
REFERENCES
(1) Bottaro, S.; Banás,̌ P.; Sponer, J.; Bussi, G. J. Phys. Chem. Lett. 2016, 7, 4032−4038.
© XXXX American Chemical Society
1674
DOI: 10.1021/acs.jpclett.8b00754 J. Phys. Chem. Lett. 2018, 9, 1674−1675
The Journal of Physical Chemistry Letters
Addition/Correction
Figure 5. Folding free energy as a function of the temperature obtained from MD simulations compared with the prediction obtained from the nearest neighbor (NN) model. Statistical errors, calculated using blocks of 200 ns, are on the order of 1 kJ/mol and are not shown for clarity. Statistical error is higher in the reweighted ensembles due to the lower effective sample size but still allows ΔG to be estimated with a statistical error lower than 2 kJ/mol.
Figure 3. Effect on the stability of the native fold (for UUCG/GAGA) and of the extended conformation upon addition of a Gaussian potential on α and ζ angles. Blue indicates that the additional potential stabilizes the correct structure, and red indicates that the native fold is destabilized. Results for tetranucleotides are reported for completeness but are identical to those reported in the original paper.
Figure 4. Folding free energy changes upon addition of a cosine correction f(θ) = cos(θ + 4.5) to α, ζ, and α + ζ angles for different tetraloops and tetranucleotides. ΔG (in kJ/mol) calculated using the uncorrected AmberχOL3 force field are reported in the figure. Results for tetranucleotides are reported for completeness but are identical to those reported in the original paper.
1675
DOI: 10.1021/acs.jpclett.8b00754 J. Phys. Chem. Lett. 2018, 9, 1674−1675
