Figure 1.
Searching for common RNA secondary structure in unaligned sequences.
The scenario of searching for common RNA structure in sequences (left) that are otherwise unrelated (here generated by shuffling the order of the nucleotides in real sequences). This structure can either represent portions of an ncRNA gene or a structural RNA element in an mRNA. The search result in a multiple structural alignment (right) is typically based on the pattern of obtained compensating changes.
Figure 2.
Basic flow homology (left) search in combination with identification of syntenic regions (right) of related genomes.
(Figure courtesy of Christian Anthon.)
Figure 3.
An example of 2D (left) and 3D (right) representations of RNA structures, here illustrated for a tRNA.
The RNA secondary structure is an important step towards the full 3D structure. (Figure from [116].)
Figure 4.
Representations of RNA (secondary) structure.
From top left: A circle plot, a conventional secondary structure diagram, a mountain plot, and a dot plot. The bottom diagram shows the secondary structure in dot-bracket notation, where a base pair is represented by matching parentheses. The respective colors in each diagram represent the same base pairs. The structure shown is a glycine riboswitch from B. subtilis, Rfam family RF00504.
Figure 5.
Decomposition of RNA secondary structures for the Nussinov algorithm.
The decomposition is unambiguous in the sense that each structure can only be decomposed in a single way.
Figure 6.
Free energies for stacked pairs and loops in kcal/mol.
Note that both base pairs have to be read in 5′-3′ direction.
Figure 7.
Filled dynamic programming matrix for the toy sequence AGCACACAGGC.
Values giving rise to the optimal folding energy of are shown in red.
Figure 8.
Structure prediction for two non-coding RNA sequences (DsrA and DicF) and respective (shuffled) sequences with the same length and nucleotide composition.
Most readers will not be able to distinguish between the real and randomized scenarios.
Figure 9.
Two toy sequences that, if aligned only by their sequence, do not match in secondary structure.
If correctly aligned, low sequence similarity between the two sequences does not hinder the revelation of structure.
Figure 10.
Filling out the dynamical programming matrix “ahead of time”.
That is, for the current position in the sequence just partially filling out future cells, either for the first time, or by updating the maximum score in the particular cell. All grey cells, including the blue cell and the current cell ( of a single sequence), have been completely computed. The green and yellow cells are partially filled out, making part use of the red cells (previously computed). (The figure is from the supplemental material of [81].)
Figure 11.
The basic flow of strategies for de novo prediction of RNA structures in genomic sequences.
Given the strategy of applying multiple organism sequences, orthologs are already obtained. For the homology search using the obtained de novo candidates, these can be compared in syntenic regions as for obtained homology candidates. (Figure courtesy of Christian Anthon.)
Figure 12.
Computation of the SCI from a multiple alignment.
Figure 13.
Searching unaligned sequences using CMfinder.
After construction of an initial alignment (based on energy folded seeds), a covariance model is constructed and used to make further searches. Additional findings are incorporated into the model and novel searches are made until convergence was reached. (The figure was kindly provided by Zizhen Yao.)