General information

How to cite GArna

GArna Workgroup

GArna publications

User's guide

Primary structure analysis

Secondary structure analysis


Additional information

Main principles of MatrixSS program

The program MatrixSS is designed for searching in genome sequences for the regions with potentially significant secondary structure (SS). It is assumed that SS is determined by the structure of RNA that could be translated from this region.

The MatrixSS program is based on two principles:

  1. E-score value calculated by the formula , where PA, PU, PG, PC are nucleotide concentrations in RNA sequence. This value is in a good agreement (r=-0.91) with the free energy value of SS of RNA with randomly occurring nucleotides. The more is the E-score value in a certain RNA, this molecule is expected to have the more stable potential SS.
  2. RNA molecules referring to any class of structural RNA are characterised by the similar nucleotide content that varies in a small range (under the structural RNA we understand those RNA classes that are characterised by functional significance of SS, e.g., tRNA, ribosomal RNA, small nuclear RNA).This property strictly differs from the properties of arbitrary genome sequences with a large variety in nucleotide content.

As follows from the principle 2, for the structural RNA the E-score value, like mononucleotide content, varies a little. The program MatrixSS applies this property for determining the regions with nucleotide content preferable for formation of functionally significant SS. If the E-score value in some genome DNA region is less than some threshold value

average(E-scorestruct) - standard_deviation(E-scorestruct)

or it exceeds the threshold value

average(E-scorestruct) + standard_deviation(E-scorestruct),

then the program MatrixSS considers this region as lacking potential SS. Here
average(E-scorestruct) denotes the mean E-score value of structural RNA, and
standard_deviation(E-scorestruct) is a standard deviation.

Note that too large E-score value also evidences about the fact that the sequence analysed is not referred to the class of structural RNA. However, this does not mean that this sequence is enable to form SS. On the contrary, in this case, the potential SS would be too much stable for performing any functions (this viewpoint is supported by 2 statements: (1) very stable SS is formed very slowly; (2) the form of very stable SS is very susceptible to mutations, hence it is not selected in course of evolution). The exception is 5'UTR sequences in eukaryotes. It is known that highly stable SS of 5'UTR RNA is capable to inhibit its translation, that is, this high stability is of biological significance. However, this function is independent from the form of SS, but the energy (stability) of SS is significant. Taking this into account, eukaryotic mRNAs (in particular, their 5'UTR regions) can not be referred to the class of structural RNA.


[an error occurred while processing this directive]