Table 1.
Benchmark results.
Figure 1.
Spontaneous cleavage pattern resulting from in-line probing of yeast asp-tRNA, nucleotides with larger backbone flexibility will have higher rates of cleavage and thus bands of greater intensity. Lanes for no reaction, T1 RNase (cleavage following only guanosines), and partial hydroxyl cleavage (-OH, cleavage after each base) are indicated. Due to the high resolution of the gel, double bands appear for nucleotides 2–9. These bands correspond to RNA molecules where the cyclic phosphate intermediate has hydrolyzed to leave either no phosphate, or a mixture of
- and
-phosphate products which migrate more quickly on the gel. Quantifcation of these positions combined the bands corresponding to both products. The precursor RNA and T1 RNase cleavage products are marked. Not all guanosines show cleavage due to retention of secondary structure at 5 M urea and elevated temperature.
Figure 2.
Normalized (blue circles) and raw (red diamonds) shape values. Gray bars indicate the missing shape values. The subplots shows the piecewise normalization map.
Figure 3.
Distribution of shape discrepancies in yeast asp-tRNA (top) and E. coli phe-tRNA (bottom). shape data for asp-tRNA [resp. phe-tRNA] from the Weeks Lab [25] [resp. Das Lab [26]]. Using crystal structure as ‘gold standard’, red squares indicate locations where the absolute value of the difference of shape data and crystal structure (1 unpaired, 0 paired) exceeds 0.5. The plots on the right show the distribution of the discrepancy in shape as well as the error rate.
Figure 4.
Comparison of In-line probing and shape.
Distribution of reactivities of data from in-line probing (A) and shape (B). In-line probing reactivities were determined using SAFA [24] and then normalized to range , in order to be comparable with shape reactivities. Histograms suggest that in-line probing signal is more diffuse than that from shape. The fraction of base-pairs in asp-tRNA is
which could be used to estimate the threshold shape moderate reactivity.
Figure 5.
The plots show heat maps displaying ppv () as a function of parameter
for RNAsc with data from shape and in-line probing (asp-tRNA
). Note the much larger area for good parameter choices when using shape data, rather than in-line probing data. This data suggests that shape data is more robust than in-line probing data, when used in computing MFE structure with RNAsc. Computations were done at 37
C.
Figure 6.
Heat maps of in-line probing and shape.
Heat maps illustrating differences between in-line probing (left) and shape (right) analysis of the yeast asp-tRNA. Nucleotides are colored corresponding to cumulative activities described in Figure 3, where the least reactive of bases are black (
of bases are paired in the crystal structure), the most reactive
of bases are red, and the next most reactive
are yellow. Gray bases are bases for which there is no data available.
Figure 7.
Pointwise entropy of yeast asp-tRNA, computed from RNAsc using shape data (red squares), in-line probing (blue diamonds), and using no probing data (black circles). Average pointwise entropies: 0.210 (shape data), 0.267 (in-line probing), 0.269 (no data). As expected, by integrating either shape or in-line probing data into RNAsc, the variability (entropy) decreases; however, it appears that variability (entropy) is decreased more by shape than by in-line probing data – again, suggesting that shape data is more robust than in-line probing data when used with RNAsc.
Figure 8.
Expected distance of predicted probabilities with normalized shape data.
The figure shows a plot of the expected distance between normalized experimental shape values
and the low energy Boltzmann ensemble, as computed by RNAsc. The
-axis depicts increasing values of RNAsc parameter
, while the
-axis depicts expected distance
. The curves confirm the statement of Theorem 2, which states that as
increases, the expected distance
decreases. The figure also shows that for higher values of
,
can be made to agree very closely
. The expected distances of the predicted probabilities with unnormalized shape values for RNAstructure are
,
, and
for asp-tRNA, HCV, and P546 respectively using optimal parameter values (
and
).
Figure 9.
Errors in the prediction of the secondary structure of glycine riboswitch by RNAsc.
On the -axis, nucleotide positions are displayed, where the algorithm predicts the structure incorrectly. The
-axis represents the shape distance to the native structure at the given nucleotide. A shape distance with absolute value
indicates an error.