Bulletin of Mathematical Biology

, Volume 59, Issue 2, pp 339–397

Generic properties of combinatory maps: Neutral networks of RNA secondary structures

  • Christian Reidys
  • Peter F. Stadler
  • Peter Schuster
Article

DOI: 10.1007/BF02462007

Cite this article as:
Reidys, C., Stadler, P.F. & Schuster, P. Bltn Mathcal Biology (1997) 59: 339. doi:10.1007/BF02462007

Abstract

Random graph theory is used to model and analyse the relationship between sequences and secondary structures of RNA molecules, which are understood as mappings from sequence space into shape space. These maps are non-invertible since there are always many orders of magnitude more sequences than structures. Sequences folding into identical structures formneutral networks. A neutral network is embedded in the set of sequences that arecompatible with the given structure. Networks are modeled as graphs and constructed by random choice of vertices from the space of compatible sequences. The theory characterizes neutral networks by the mean fraction of neutral neighbors (λ). The networks are connected and percolate sequence space if the fraction of neutral nearest neighbors exceeds a threshold value (λ>λ*). Below threshold (λ<λ*), the networks are partitioned into a largest “giant” component and several smaller components. Structure are classified as “common” or “rare” according to the sizes of their pre-images, i.e. according to the fractions of sequences folding into them. The neutral networks of any pair of two different common structures almost touch each other, and, as expressed by the conjecture ofshape space covering sequences folding into almost all common structures, can be found in a small ball of an arbitrary location in sequence space. The results from random graph theory are compared to data obtained by folding large samples of RNA sequences. Differences are explained in terms of specific features of RNA molecular structures.

Nomenclature

v[G]

Vertex set of graphG

e[G]

Edge set of graphG

ω(X)

Cardinality ofX as a set

δv

Vertex degree in a corresponding graphG

Qαn

Generalized hypercube

\(\hat X\)

X is a random variable

E[\(\hat X\)]

Expectation value of the random variable\(\hat X\)

V[\(\hat X\)]

Variance of\(\hat X\)

E[\(\hat X\)]r

rth factorial moment of\(\hat X\)

μn, λ,μn

Measure\(\mu _n (\Gamma _n )\mathop = \limits^{def} \lambda ^{\omega (v[\Gamma ])} (1 - \lambda )^{\omega (v[H]) - \omega (v[\Gamma ])} \)

Ωn

Probability space ({Γn},μn, λ

\(\hat X_{n,k} \)

Number of vertices in a random graph Γn having degreek

\(\hat I_n (\Gamma _n )\)

=ω({v∈v[Γn]|∂{v}∩v[Γn]=∅}), i.e., the number of isolated vertices in a random graph Γn

\(\hat Z_n (\Gamma _n )\)
i.e. the number of vertices inQαn that are at least of distance 2 w.r.t. a random graph Γn
\(M_{n,k}^{\upsilon ,\upsilon '} (\Gamma _n )\)

Set of paths {π1)|π1)∈Π(Γn)}

\(\hat Y_{n,k}^{\upsilon ,\upsilon '} \)

\( = \omega (M_{n,k}^{\upsilon ,\upsilon '} (\Gamma _n ))\) and 0 otherwise

\(\hat \Lambda _{n,k} \)

Random variable that is 1 if all pairs υ,υ′∈v[Γn] withd(v, v′)<k occur in a path ofd(υ,υ′)<k and 0 otherwise

GV

Set of adjacent vertices w.r.t. a vertex setV⊂v[G] in a graphG

\(\bar V\)

v[V]∪∂V, i.e. the closure ofV

r(v)
the “ball” with radiusr and centerv
n

Chain length

nu,np

Number of unpaired and paired based of a certain secondary structure

γn

(α−1)n, i.e. the vertex degrees ofQαn

s

RNA secondary structure inn vertices

Π(s

\(\mathop = \limits^{def} \left\{ {\left[ {i,k} \right]|a_{i,k} = 1,k \ne i - 1,i + 1} \right\}\) i.e. the set of contacts of the secondary structures

Ln

Shape space, in particular, the space of RNA secondary structures inn vertices

C[s]

Graph of compatible sequences with respect tos

C[s]

v[C[s]], the set of compatible sequences

Sn

Permutation group ofn letters

Dm

Dihedral group of order2m

ΦxΓ

G1×G2{y∈v[G2]⋎(x, y)∈v[⩾s]}], the fiber inx

ΦyΓ

G1×G2[{x∈v[G1]⋎(x, y)∈v[⩽s]}], the fiber iny

ΓnA[s]
Random induced subgraph of
according to model A
ΓnB[s]
Random induced subgraph of
according to model B
dist(Γ1, Γ2)

Minimum Hamming distance between the graph Γ1 and Γ2 considered as subgraph ofQαn

Copyright information

© Society for Mathematical Biology 1997

Authors and Affiliations

  • Christian Reidys
    • 1
    • 2
  • Peter F. Stadler
    • 1
    • 2
  • Peter Schuster
    • 1
    • 2
    • 3
    • 4
  1. 1.Santa Fe InstituteSanta FeU.S.A.
  2. 2.Los Alamos National LaboratoryLos AlamosU.S.A.
  3. 3.Institut für Theoretische Chemie der Universität WienWienAustria
  4. 4.Institut für Molekulare BiotechnologieJenaGermany