Sykes and Levitt 2005 JMB

First they define a doublet as an object where any 2 heavy atoms in nucleotides i and i+1 are less that 4 Angstrom apart, the heavy atoms used are not all but just 3 in the base and 3 in the backbone. Then they use k-means (with RMSD as the metric) and to be able to specify the number of clusters (n) that they want to come out of the data, they vary n as 10, 20, 30 40, 50 and 100. Of course the clusters are just sets of data so these sets are given the name of libraries. The cut off for X-ray resolution is 4 Angstrom

Lucas and Dill 2003 JCP

Two problems are the main ones in RNA theory

1)Predict native structure based on its nucleotide sequence.
2)Predict folding thermodynamics.

The second point is the one that Dill's group has mainly focused its attention using a statistical thermodynamics model coupled to secondary structure graphs which are treated using so called polymer graphs. For treating pseudoknots this are split into "pseudoknot core units" and then determine the number of conformations for this core units.

Henri Orland and A. Zee, 2002, Nuc. Phys. B

In contrast to the problem of protein folding, RNA folding is hierarchical in that its secondary structure is much more stable than its tertiary structure, which can be treated as a perturbation. Experimentally, the two levels of folding (secondary and tertiary) can be separated by varying the concentration of Mg++ ions.


RNA Folding in vitro and in vivo can be comparable in some situation but not in others. It seems that RNA folding can be facilitated in vivo in ocassions where it does not occur in vitro as in pre-rRNA expression in e.coli. in vitro and in vivo scientists (as opposed to "in silico" and "in cogito") propose two models for large RNA molecule folding, that is;

1) Random Coil forms after transcription, non-directional folding follows (seen in vitro)
2) Folding occurs during transcription as directed by sequence (proposed for in vivo)

3 Facts of transcription influence the folding process:

1) Directionality.
2) Velocity.
3) Pausing.

Divalent ions are also mentiones as crucial both in vivo and in vivo specially in relation to splicing events. In prokaryotes transcription and translation are coupled. RNA chaperones can accelerate folding. ncRNA's can modulate transcription. It is also said that as of 2002 in vivo studies of folding is in its begginings and "lacks adquate experimental methods".

Schlick 2002 in Molecular Modeling and Simulation

3 Levels in nucleic acids:
Sugar == puckers, also torsion angles
Phosphate backbone == torsion angles as in schneider bergman
Base == there's where the bonding interactions are at, H-bonds and stacking.

From Westhof and Fritsch 2000 Structure

There are diverse base pairs which are different from traditional, Watson and Crick or Hooogsteen alone, for example one G.A pair mediated by water to make hydrogen bonds. They reinforce the concept of the 3 faces of bases, that is, W-C edge, Hoogsteeen edge and Shallow groove edge, it points out that H-bonding in the shallow groove is common. Also interstrand and cross-strand stacking. Remarks W-C form helical framework and points out non W-C are primordial in RNA-RNA and RNA-protein interaction.
RNA Tertiary (RNA-RNA interactions):
==> Between helices.
==> Between unpaired regions.
==> Between unpaired and double standed helix.