1:
2:
3: %%%%%%%%%%%%%%%%%%%%%%%% Ams-Style %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
4: %%%
5: %%% Style and Inputs
6: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
7:
8:
9: \documentclass[10pt]{amsart}
10: \usepackage{amssymb}
11: %\usepackage{CJK,CJKnumb}
12: \usepackage{amsmath,amssymb,amsfonts,amsthm,
13: latexsym, amscd, amsfonts, epsfig}
14: \usepackage{mathrsfs}
15: \usepackage{color}
16: \usepackage{eucal}% caligraphic-euler fonts: \mathcal{ }
17: \usepackage{eufrak}% frak-euler fonts: \mathfrak{ }
18: \usepackage[all]{xypic}
19: \usepackage{xspace}
20:
21: %%%
22: %%%
23: %%%%%%%%%%%%%%%%%%%%%%%%% Pagestyle %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
24: %%%
25: %%%
26:
27: \renewcommand{\baselinestretch}{1.2}% spacing between lines
28:
29: %\hoffset=0truecm
30: %\voffset=0truecm
31: \textwidth=15truecm
32: \textheight=18truecm
33: \baselineskip=0.8truecm
34: \overfullrule=0pt
35: \parskip=0.8\baselineskip
36: \parindent=0truecm
37: \topmargin=0.5truecm
38: \headsep=1.2truecm
39: %\oddsidemargin=0.5in % options for double-side printouts
40: %\evensidemargin=0in
41:
42: %%%
43: %%%
44: %%%%%%%%%%%%%%%%%%%% New Settings %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
45: %%%
46: %%%
47:
48: \theoremstyle{plain}
49: \newtheorem{theorem}{Theorem}
50: \newtheorem{corollary}{Corollary}
51: \newtheorem*{main}{Main~Theorem}
52: \newtheorem{lemma}{Lemma}
53: \newtheorem{proposition}{Proposition}
54: \newtheorem{remark}{Remark}
55: \theoremstyle{definition}
56: \newtheorem{definition}{Definition}
57:
58: \theoremstyle{example}
59: \newtheorem{example}{Example}
60:
61: \theoremstyle{remark}
62:
63: \numberwithin{equation}{section}
64:
65: % references alphabetically
66: % make relation clearer
67:
68: \begin{document}
69:
70: %%%
71: %%%
72: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
73: %%
74: %%%
75: \title[Pseudoknot RNA Structures with Arc-Length $\ge 3$]
76: {Pseudoknot RNA Structures with Arc-Length $\ge 3$}
77: \author{Emma Y. Jin and Christian M. Reidys$^{\,\star}$}
78: \address{Center for Combinatorics, LPMC-TJKLC \\
79: Nankai University \\
80: Tianjin 300071\\
81: P.R.~China\\
82: Phone: *86-22-2350-6800\\
83: Fax: *86-22-2350-9272}
84: \email{reidys@nankai.edu.cn}
85: \thanks{}
86: \keywords{Asymptotic enumeration, RNA secondary structure, $k$-noncrossing
87: RNA structure, pseudoknot, generating function, transfer theorem,
88: Hankel contour, singular expansion}
89: \date{June 2007}
90: \begin{abstract}
91: In this paper we study $k$-noncrossing RNA structures with
92: arc-length $\ge 3$, i.e.~RNA molecules in which for any $i$, the
93: nucleotides labeled $i$ and $i+j$ ($j=1,2$) cannot form a bond and
94: in which there are at most $k-1$ mutually crossing arcs. Let ${\sf
95: S}_{k,3}(n)$ denote their number. Based on a novel functional
96: equation for the generating function $\sum_{n\ge 0}{\sf
97: S}_{k,3}(n)z^n$, we derive for arbitrary $k\ge 3$ exponential growth
98: factors and for $k=3$ the subexponential factor. Our main result is
99: the derivation of the formula ${\sf S}_{3,3}(n) \sim
100: \frac{6.11170\cdot 4!}{n(n-1)\dots(n-4)} 4.54920^n$.
101: \end{abstract}
102: \maketitle
103: {{\small
104: %\tableofcontents
105: }}
106:
107:
108:
109: %%%
110: %%%
111: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
112: %%%
113: %%%
114:
115: \section{Introduction}
116:
117: %%%
118: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
119: %%%
120:
121: RNA is believed to be central for the understanding of evolution.
122: It acts as genotypic legislative in form of viruses and viroids and
123: as phenotypic executive in form of ribozymes, capable of catalytic
124: activity, cleaving other RNA molecules.
125: This dualism gives rise to the hypothesis that RNA may have preceded
126: DNA and proteins, therefore playing a key role in prebiotic evolution.
127: In light of growing support of an {\it RNA world} \cite{Schuster:02} and
128: even RNA-based metabolisms or the prospect of self-replicating RNA
129: \cite{Poole:06a} it is the phenotypic aspect of RNA that still lacks
130: deeper understanding.
131: Despite the fact that pseudoknot RNA structures are known to be of central
132: importance \cite{Science:05a}, little is known from a theoretical
133: point of view, for instance only recently their generating function has been
134: obtained \cite{Reidys:07pseu}.
135:
136: Let us provide first some background on RNA sequences and structures which
137: allows us to put our results into context.
138: The primary sequence of an RNA molecule is the sequence of
139: nucleotides {\bf A}, {\bf G}, {\bf U} and {\bf C} together with the
140: Watson-Crick ({\bf A-U},{\bf G-C}) and ({\bf U-G}) base pairing
141: rules specifying the pairs of nucleotides can potentially form
142: bonds. Single stranded RNA molecules form helical structures whose
143: bonds satisfy the above base pairing rules and which, in many cases,
144: determine or are even tantamount to their function.
145: Three decades ago Waterman {\it et.al.} pioneered the concept
146: of RNA secondary structures
147: \cite{Penner:93c,Waterman:79a,Waterman:78a,Waterman:94a,Waterman:80},
148: a concept being best understood when considering structure as a diagram,
149: drawing the primary sequence of nucleotides
150: horizontally and ignoring all chemical bonds of its backbone. Then one
151: draws all bonds, i.e.~nucleotide interactions satisfying the Watson-Crick
152: base pairing rules (and {\bf G}-{\bf U} pairs) as arcs in the upper
153: halfplane, thereby identifying structure with the set of all
154: arcs. Then secondary structures have no two arcs $(i_1,j_1)$, $(i_2,j_2)$,
155: where $i_1<j_1$ and $i_2<j_2$ with the property $i_1<i_2<j_1<j_2$ and
156: all arcs have at least length $2$.
157: While the concept of secondary structure is of fundamental
158: importance it is well-known that there exist additional types of
159: nucleotide interactions \cite{Science:05a}. These bonds are called
160: pseudoknots \cite{Westhof:92a} and occur in functional RNA (RNAseP
161: \cite{Loria:96a}), ribosomal RNA \cite{Konings:95a} and are
162: conserved in the catalytic core of group I introns.
163: In plant viral RNAs pseudoknots mimic tRNA structure and in {\it in
164: vitro} RNA evolution \cite{Tuerk:92} experiments have produced
165: families of RNA structures with pseudoknot motifs, when binding
166: HIV-1 reverse transcriptase.
167:
168: Leaving the paradigm of RNA secondary structures, i.e.~studying RNA structures
169: with crossing bonds, poses challenging problems for computational biology.
170: Prediction algorithms for RNA pseudoknot structures are much harder
171: to derive since there exists no {\it a priori} tree-structure and
172: the subadditivity of local solutions is not guaranteed. In fact pseudoknot
173: RNA structures are genuinely non-inductive and seem to be best described by the
174: mathematical language of vacillating tableaux \cite{Reidys:07pseu,Chen:07a}.
175: One approach of categorizing RNA pseudoknot structures consists in considering
176: the maximal size of sets of mutually crossing bonds,
177: leading to the notion of $k$-noncrossing RNA structures \cite{Reidys:07pseu}.
178: %%%
179: %%%%%%%%%%%%%%%%%%%%%%%% Figures ex1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
180: %%%
181: \begin{figure}[ht]
182: \centerline{%
183: \epsfig{file=f1.eps,width=0.7\textwidth}\hskip15pt } \caption{\small
184: $k$-noncrossing RNA structures. (a) secondary structure (with
185: isolated labels $3,7,8,10$), (b) non-planar $3$-noncrossing
186: structure } \label{F:4}
187: \end{figure}
188: %%%
189: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
190: %%%
191: This concept is very intuitive, a $k$-noncrossing RNA structure has at most
192: $k-1$ mutually crossing arcs and a minimum bond-length of $2$, i.e.~for any
193: $i$, the nucleotides $i$ and $i+1$ cannot form a bond.
194: In this paper we will consider $k$-noncrossing RNA structures with arc-length
195: $\ge 3$. Their analysis is based on the generating function, which
196: has coefficients that are alternating sums. This fact makes even the
197: computation of the exponential growth factor a difficult task.
198: To make things worse, in case of arc-length $\ge 3$ there exists no
199: explicit formula for the coefficients, which could only be computed
200: via a recursion formula.
201:
202: \subsection{Organization and main results}
203: Let $\mathcal{S}_{k,3}(n)$ ($k\ge 3)$ denote the set of $k$-noncrossing RNA
204: structures with arc-length $\ge 3$ and let ${\sf S}_{k,3}(n)= \vert
205: \mathcal{S}_{k,3}(n)\vert $. In Section~\ref{S:pre} we provide the
206: necessary background on $k$-noncrossing RNA structures with
207: arc-length $\ge 3$ and their generating function $\sum_{n\ge 0}{\sf
208: S}_{k,3}(n)z^n$. In Section~\ref{S:exp} we compute the exponential
209: factor of ${\sf S}_{k,3}(n)$. To make it easily accessible to a broad
210: readership we give an elementary proof based on real analysis and
211: transformations of the generating function.
212: Central to our proof is the functional identity of
213: Lemma~\ref{L:func} and its generalized version in Lemma~\ref{L:ana}.
214: In Section~\ref{S:sub} we present the asymptotic analysis of
215: $\mathcal{S}_{3,3}(n)$, using Flajolet {\it et.al.}'s singular
216: expansions and transfer theorems
217: \cite{Flajolet:05,Flajolet:99,Flajolet:94,Popken:53,Odlyzko:92}.
218: This analysis is similar to \cite{Reidys:07pseu} but involves solving a
219: quartic instead of a quadratic polynomial in order to localize the
220: singularities. The main result of the paper is
221:
222: %%%
223: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
224: %%%
225: {\bf Theorem.}$\,$ {\it The number of $3$-noncrossing RNA structures
226: with arc length $\ge 3$ is asymptotically given by
227: \begin{eqnarray}
228: \label{E:konk3} {\sf S}_{3,3}(n) & \sim & \frac{6.11170\cdot
229: 4!}{n(n-1)\dots(n-4)}\,
230: 4.54920^n \ ,
231: \end{eqnarray}
232: where ${\sf s}_{3,3}^{}(n)=\frac{6.11170\cdot 4!}{n(n-1)\dots(n-4)}=
233: 146.6807\left[\frac{1}{ n^5}-\frac{35}{4n^6}+\frac{1525}{32
234: n^7}+{O}(n^{-8})\right]$.
235: }
236: %%%
237: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
238: %%%
239:
240: In the table below we display the quality of our approximation by listing
241: the subexponential factors, i.e.~we compare for $k=3$ the quantities
242: ${\sf S}_{k,3}(n)/(4.54920)^n$
243: obtained from the generating function (Theorem~\ref{T:cool2}), which
244: are the exact values and the asymptotic expressions ${\sf
245: s}_{3,3}^{}(n)$, respectively.
246: \begin{center}
247: \begin{tabular}{c|c|c|c|c|c}
248: \hline
249: \multicolumn{6}{c}{\textbf{The sub exponential factor}}\\
250: \hline
251: $n$ & ${\sf S}_{3,3}(n)/(4.54920)^n$ & ${\sf s}_{3,3}^{}(n)$ &$n$ & ${\sf S}_{3,3}(n)/(4.54920)^n$ & ${\sf s}_{3,3}^{}(n)$\\
252: \hline \small 10 & \small $3.016\times 10^{-4}$ & \small$4.851\times 10^{-3}$ &
253: \small 60 & \small $3.457 \times 10^{-7}$ & \small$2.238\times 10^{-7}$\\
254: \small 20 & \small $2.017 \times 10^{-5}$ & \small$7.884\times 10^{-5}$ &
255: \small 70 & \small $1.476\times 10^{-7}$ & \small$1.010\times 10^{-7}$\\
256: \small 30 & \small $3.513 \times 10^{-6}$ & \small$8.577\times
257: 10^{-6}$ & \small 80 & \small $3.783\times 10^{-8}$ &
258: \small$5.085\times
259: 10^{-8}$\\
260: \small 40 & \small $9.646\times 10^{-7}$ & \small$1.858\times
261: 10^{-6}$ & \small 90 & \small $2.154\times 10^{-8}$ & \small$2.781\times 10^{-8}$\\
262: \small 50 & \small $5.627\times 10^{-7}$ & \small$5.769\times
263: 10^{-7}$ &
264: \small 100 & \small $1.299\times 10^{-8}$ & \small$1.624\times 10^{-8}$\\
265: \end{tabular}
266: \end{center}
267: %%%
268: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
269: %%%
270:
271: \section{$k$-noncrossing RNA structures with arc-length $\ge 3$}\label{S:pre}
272:
273: %%%
274: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
275: %%%
276: Suppose we are given the primary RNA sequence
277: $$
278: {\bf A}{\bf C}{\bf U}{\bf C}{\bf A}{\bf G}{\bf U}{\bf U}{\bf A} {\bf
279: G}{\bf A}{\bf A}{\bf U}{\bf A}{\bf G}{\bf C}{\bf C}{\bf G}{\bf G}
280: {\bf U}{\bf C} \ .
281: $$
282: We then identify an RNA structure with the set of all bonds
283: different from the backbone-bonds of its primary sequence, i.e.~the
284: arcs $(i,i+1)$ for $1\le i\le n-1$. Accordingly an RNA structure is
285: a combinatorial graph over the labels of the nucleotides of the
286: primary sequence. These graphs can be represented in several ways.
287: In Figure~\ref{F:2} we represent a structure with loop-loop
288: interactions in two ways.
289: %%%%
290: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
291: %%%%
292: \begin{figure}[ht]\label{F:2}
293: \centerline{ \epsfig{file=f2.eps,width=0.8\textwidth}\hskip15pt }
294: \caption{\small A $3$-noncrossing RNA structure with arc-length $\ge
295: 3$, as a planar graph (top) and as a diagram (bottom)} \label{F:6}
296: \end{figure}
297: %%%%
298: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
299: %%%%
300: In the following we will consider structures as diagram
301: representations of digraphs. A digraph $D_n$ is a pair of sets
302: $V_{D_n},E_{D_n}$, where $V_{D_n}=
303: \{1,\dots,n\}$ and $E_{D_n}\subset \{(i,j)\mid 1\le i< j\le n\}$.
304: $V_{D_n}$ and $E_{D_n}$ are called vertex and arc set, respectively.
305: A $k$-noncrossing digraph is a digraph in which all
306: vertices have degree $\le 1$ and which does not contain a $k$-set of
307: arcs that are mutually intersecting, or equivalently
308: \begin{eqnarray}
309: \ \not\exists\,
310: (i_{r_1},j_{r_1}),(i_{r_2},j_{r_2}),\dots,(i_{r_k},j_{r_k});\quad & &
311: i_{r_1}<i_{r_2}<\dots<i_{r_k}<j_{r_1}<j_{r_2}<\dots<j_{r_k} \ .
312: \end{eqnarray}
313: We will represent digraphs as a diagrams (Figure~\ref{F:2}) by
314: representing the vertices as integers on a line and connecting any
315: two adjacent vertices by an arc in the upper-half plane. The
316: direction of the arcs is implicit in the linear ordering of the
317: vertices and accordingly omitted.
318: %%%
319: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
320: %%%
321: \begin{definition}\label{D:rna}
322: An $k$-noncrossing RNA structure with arc-length $\ge 3$ is a digraph in
323: which all vertices have degree $\le 1$, having at most a $k-1$-set of
324: mutually intersecting arcs without arcs of length $\le 2$, i.e.~arcs of
325: the form $(i,i+1)$ and $(i,i+2)$, respectively.
326: Let ${\sf S}_{k,3}(n)$ and ${\sf S}_{k,3}(n,\ell)$ be the numbers of
327: $k$-noncrossing RNA structures with arc-length $\ge 3$ and those with
328: $\ell$ isolated vertices, respectively.
329: \end{definition}
330: %%%
331: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
332: Let $f_{k}(n,\ell)$ denote the number of $k$-noncrossing digraphs with
333: $\ell$ isolated points. We have shown in \cite{Reidys:07pseu} that
334: \begin{align}\label{E:ww0}
335: f_{k}(n,\ell)& ={n \choose \ell} f_{k}(n-\ell,0) \\
336: \label{E:ww1}
337: \det[I_{i-j}(2x)-I_{i+j}(2x)]|_{i,j=1}^{k-1} &=
338: \sum_{n\ge 1} f_{k}(n,0)\,\frac{x^{n}}{n!} \\
339: \label{E:ww2}
340: e^{x}\det[I_{i-j}(2x)-I_{i+j}(2x)]|_{i,j=1}^{k-1}
341: &=(\sum_{\ell \ge 0}\frac{x^{\ell}}{\ell!})(\sum_{n \ge
342: 1}f_{k}(n,0)\frac{x^{n}}{n!})=\sum_{n\ge 1}
343: \left\{\sum_{\ell=0}^nf_{k}(n,\ell)\right\}\,\frac{x^{n}}{n!} \ .
344: \end{align}
345: where $I_{r}(2x)=\sum_{j \ge 0}\frac{x^{2j+r}}{{j!(r+j)!}}$ is the
346: hyperbolic Bessel function of the first kind of order $r$. In
347: particular we have for $k=3$
348: \begin{equation}\label{E:2-3}
349: f_{3}(n,\ell)=
350: {n \choose \ell}\left[C_{\frac{n-\ell}{2}+2}C_{\frac{n-\ell}{2}}-
351: C_{\frac{n-\ell}{2}+1}^{2}\right] \ ,
352: \end{equation}
353: where $C_m=\frac{1}{m+1}\binom{2m}{m}$ is the $m$th Catalan number.
354: The derivation of the generating function of $k$-noncrossing RNA structures,
355: given in Theorem~\ref{T:cool2} below uses advanced methods and novel
356: constructions of enumerative combinatorics due to Chen~{\it et.al.}
357: \cite{Chen:07a,Gessel:92a} and Stanley's mapping between matchings and
358: oscillating tableaux i.e.~families of Young diagrams in which any two
359: consecutive shapes differ by exactly one square.
360: The enumeration is obtained using the reflection principle due to Gessel
361: and Zeilberger \cite{Gessel:92a} and Lindstr\"om \cite{Lindstroem:73a}
362: in accord with an inclusion-exclusion argument in order to eliminate the
363: arcs of length $\le 2$ \cite{Reidys:07pseu}.
364: %%%
365: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
366: %%%
367: \begin{theorem}\label{T:cool2}
368: Let $k\in\mathbb{N}$, $k>2$. Then the numbers of $k$-noncrossing
369: RNA structures ${\sf S}_{k,3}(n,\ell)$ and ${\sf S}_{k,3}(n)$ are given by
370: \begin{eqnarray}\label{E:da2}
371: {\sf S}_{k,3}(n,\ell) & = &
372: \sum_{b\ge 0}(-1)^{b}
373: \lambda(n,b)f_{k}(n-2b),\ell) \\
374: \label{E:da3}
375: {\sf S}_{k,3}(n) & = & \sum_{b=0}^{\lfloor n/2\rfloor}
376: (-1)^{b} \lambda(n,b)
377: \left\{\sum_{\ell=0}^{n-2b}f_{k}(n-2b,\ell)\right\}\ .
378: \end{eqnarray}
379: where $\lambda(n,b)$ satisfies the recursion
380: \begin{equation}\label{E:hh}
381: \lambda(n,b)=\lambda(n-1,b)+\lambda(n-2,b-1)+
382: \lambda(n-3,b-1)+\lambda(n-4,b-2)
383: \end{equation}
384: and the initial conditions for eq.~{\rm (\ref{E:hh})} are
385: $\lambda(n,0)=1$ and $\lambda(n,1)=2n-3$.
386: \end{theorem}
387: %%%
388: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
389: %%%
390: %%%
391: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
392: %%%
393:
394: \section{The exponential factor}\label{S:exp}
395:
396: %%%
397: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
398: %%%
399: In this section we obtain the exponential growth factor of the
400: coefficients ${\sf S}_{k,3}(n)$. Let us begin by considering the
401: generating function $\sum_{n\ge 0}{\sf S}_{k,3}(n)x^n$ as a power series
402: over $\mathbb{R}$. Since $\sum_{n\ge 0}{\sf S}_k(n)x^n$ has
403: monotonously increasing coefficients $\lim_{n\to\infty}{\sf
404: S}_{k,3}(n)^{\frac{1}{n}}$ exists and determines via Hadamard's formula
405: its radius of convergence. Due to the inclusion-exclusion form of the
406: terms ${\sf S}_{k,3}(n)$, it is not obvious however, how to compute this
407: radius. Our strategy consists in first showing that ${\sf S}_{k,3}(n)$ is
408: closely related to $f_k(2n,0)$ via a functional relation of generating
409: functions.
410: %%%
411: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
412: %%%
413: \begin{lemma}\label{L:laplace}
414: Let $x$ be an indeterminate over $\mathbb{R}$ and $T_{k}(n)$ be the number of
415: $k$-noncrossing partial matchings over $[n]$, i.e. $T_{k}(n)=\sum_{m \le
416: \lfloor \frac{n}{2}\rfloor}\binom{n}{2m}f_k(2m,0)$.
417: Let furthermore $\rho_k$ denote the
418: radius of convergence of $\sum_{n \ge 0}T_{k}(n)\, x^n$.
419: Then we have
420: \begin{equation}\label{E:laplace}
421: \forall \, \vert x\vert <\rho_k;\quad
422: \sum_{n \ge 0}T_{k}(n)\, x^n =
423: \frac{1}{1-x}\, \sum_{n \ge 0}f_k(2n,0)\,\left(\frac{x}{1-x}\right)^{2n} \ .
424: \end{equation}
425: \end{lemma}
426: %%%
427: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
428: %%%
429: \begin{proof}
430: We will relate the ordinary generating functions of $T_{k}(n)$
431: \begin{equation}\label{E:ll}
432: \sum_{n \ge 0}T_{k}(n)\frac{x^n}{n!}=
433: \sum_{n \ge 0}\sum_{m\le\frac{n}{2}}{n \choose 2m}f_k(2m,0)
434: \frac{x^n}{n!}=e^x\cdot {\sf det}[I_{i-j}(2x)-I_{i+j}(2x)]_{i,j=1}^{k-1}\
435: \end{equation}
436: (see eq.~(\ref{E:ww2}))
437: and $f_k(2n,0)$ via Laplace transforms as follows: $\sum_{n \ge 0}
438: T_{k}(n)x^n = \sum_{n \ge 0}T_{k}(n)\frac{x^n}{n!}n!$ is convergent for
439: any $x\in\mathbb{R}$, whence we derive, using the Laplace transformation
440: and interchanging integration and summation
441: \begin{eqnarray*}
442: \sum_{n \ge 0}T_{k}(n)x^n = \sum_{n \ge 0}T_{k}(n)\frac{x^n}{n!}
443: \int_{0}^{\infty}e^{-t}t^n dt =
444: \int_0^{\infty}\sum_{n \ge 0}T_{k}(n)\frac{(xt)^n}{n!}e^{-t}dt
445: \end{eqnarray*}
446: We next interpret the rhs via eq.~(\ref{E:ll}) and obtain
447: \begin{eqnarray*}
448: \sum_{n \ge 0}T_{k}(n)x^n
449: &=& \int_0^{\infty}e^{xt}{\sf det}[I_{i-j}(2xt)-I_{i+j}(2xt)]_{i,j=1}^{k-1}
450: e^{-t}dt \ .
451: \end{eqnarray*}
452: Interchanging integration and summation yields
453: \begin{eqnarray*}
454: \sum_{n \ge 0}T_{k}(n)x^n
455: &=& \int_0^{\infty}e^{(x-1)t}\sum_{n \ge 0}f_k(2n,0)\frac{(xt)^{2n}}{(2n)!}dt\\
456: &=& \sum_{n \ge 0}f_k(2n,0)\cdot\frac{1}{(2n)!}
457: \int_0^{\infty}e^{(x-1)t}(xt)^{2n}dt\\
458: &=& \sum_{n \ge 0}f_k(2n,0)\frac{1}{(2n)!}\int_0^{\infty}
459: e^{-(1-x)t}((1-x)t)^{2n}(\frac{x}{1-x})^{2n}dt\\
460: &=&\sum_{n \ge
461: 0}f_k(2n,0)\frac{1}{(2n)!}\cdot(2n)!(\frac{x}{1-x})^{2n}\cdot\frac{1}{1-x}\\
462: &=&\frac{1}{1-x}\sum_{n \ge 0}f_k(2n,0)\,(\frac{x}{1-x})^{2n}
463: \end{eqnarray*}
464: and the proof of the lemma is complete.
465: \end{proof}
466: %%%
467: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
468: %%%
469: \begin{lemma}\label{L:func}
470: Let $x$ be an indeterminante over $\mathbb{R}$ and ${\sf S}_{k,3}(n)$
471: be the number of $k$-noncrossing RNA structures with arc-length $\ge
472: 3$. Then we have the functional equation
473: \begin{equation}\label{E:laplace}
474: \sum_{n \ge 0}{\sf S}_{k,3}(n)\, x^n =
475: \frac{1}{1-x+x^2+x^3-x^4}\sum_{n \ge
476: 0}f_k(2n,0)\left(\frac{x-x^3}{1-x+x^2+x^3-x^4}\right)^{2n}
477: \end{equation}
478: \end{lemma}
479: %%%
480: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
481: %%%
482: \begin{proof}
483: According to Theorem~\ref{T:cool2} we have
484: \begin{eqnarray*}
485: {\sf S}_{k,3}(n) & = & \sum_{b \le \lfloor \frac{n}{2}\rfloor}
486: (-1)^b\lambda(n,b)\sum_{\ell=0}^{n-2b}f_{k}(n-2b,\ell) \\
487: & = & \sum_{b \le \lfloor \frac{n}{2}\rfloor}
488: (-1)^b\lambda(n,b)\sum_{m=2b}^{n}
489: {n-2b \choose m-2b}f_{k}(m-2b,0)
490: \end{eqnarray*}
491: Our goal is now to relate $\sum_{n \ge 0}{\sf S}_{k,3}(n)x^n$ to the
492: terms $T_k(n)$. For this purpose we derive
493: \begin{align*}
494: \sum_{n \ge 0}{\sf S}_{k,3}(n)x^n&=\sum_{n \ge 0}\sum_{2b\le n}
495: (-1)^b\lambda(n,b)\sum_{m=2b}^{n}{n-2b \choose
496: m-2b}f_k(m-2b,0)\, x^n\\
497: &=\sum_{b \ge 0}(-1)^b x^{2b}\sum_{n \ge
498: 2b}\lambda(n,b)T_{k}(n-2b)x^{n-2b}\\
499: &=\sum_{b \ge 0}(-1)^b x^{2b}\sum_{n \ge 0}\lambda(n+2b,b)T_{k}(n)\,x^n \ .\\
500: \end{align*}
501: Interchanging the summations w.r.t.~$b$ and $n$ we consequently arrive at
502: \begin{equation}\label{E:21}
503: \sum_{n \ge 0}{\sf S}_{k,3}(n)x^n = \sum_{n \ge 0}\left[\sum_{b\ge 0}(-1)^b
504: x^{2b}\lambda(n+2b,b)\right]T_{k}(n)\, x^n \ .
505: \end{equation}
506: We set $\varphi_{n}(x)=\sum_{b \ge 0}\lambda(n+2b,b)x^b$. According to
507: Theorem~\ref{T:cool2} we have the recursion formula
508: $$
509: \lambda(n+2b,b)=\lambda(n+2b-1,b)+\lambda(n+2b-2,b-1)+
510: \lambda(n+2b-3,b-1)+\lambda(n+2b-4,b-2)
511: $$
512: Multiplying with $x^b$ and taking the summation over all $b$
513: ranging from $0$ to $\lfloor n/2\rfloor$ implies the following functional
514: equation for $\varphi_{n}(x)$, $n=1,2\ldots$
515: \begin{equation}\label{E:22}
516: \varphi_{n}(x)=\varphi_{n-1}(x)+x\cdot \varphi_{n}(x)+x \cdot \varphi_{n-1}(x)+
517: x^2\varphi_{n}(x) \ .
518: \end{equation}
519: Eq.~(\ref{E:22}) is equivalent to
520: \begin{equation}\label{E:23}
521: \frac{\varphi_{n}(x)}{\varphi_{n-1}(x)}=\frac{1+x}{1-x-x^2}
522: \quad \text{\rm and }\quad \varphi_{0}(x)=\sum_{b \ge
523: 0}\lambda(2b,b)x^b=\frac{1}{1-x-x^2}
524: \end{equation}
525: since $\lambda_{b}=\lambda(2b,b)$ satisfies the recursion formula
526: $\lambda_b=\lambda_{b-1}+\lambda_{b-2}$ and the initial condition
527: $\lambda_{0}=\lambda_{1}=1$. We can conclude from this that $\lambda(2b,b)$
528: is $b$-th Fibonacci number. As a result we obtain the formula
529: \begin{equation}\label{E:24}
530: \varphi_{n}(x)=\varphi_{0}(x)\left(\frac{1+x}{1-x-x^2}\right)^n=
531: \frac{1}{1-x-x^2}
532: \left(\frac{1+x}{1-x-x^2}\right)^n \ .
533: \end{equation}
534: Substituting eq.~(\ref{E:24}) into eq.~(\ref{E:21}) we can compute
535: \begin{eqnarray*}
536: \sum_{n \ge 0}{\sf S}_{k,3}(n)x^n &= & \sum_{n \ge
537: 0}\varphi_{n}(-x^2)T_{k}(n)x^n \\
538: & = &\sum_{n \ge
539: 0}\frac{1}{1+x^2-x^4}\left(\frac{1-x^2}{1+x^2-x^4}\right)^n
540: T_{k}(n)\ x^n\\
541: & = & \frac{1}{1+x^2-x^4}
542: \sum_{n \ge
543: 0} T_{k}(n)\, \left(\frac{x-x^3}{1+x^2-x^4}\right)^n \ .
544: \end{eqnarray*}
545: Via Lemma~\ref{L:laplace} we have the following interpretation of
546: $\sum_{n \ge 0}T_{k}(n) \, x^n$
547: $$
548: \sum_{n \ge 0}T_{k}(n) \, x^n = \frac{1}{1-x}\sum_{n \ge 0}
549: f_k(2n,0)\,(\frac{x}{1-x})^{2n}\ .
550: $$
551: Therefore we obtain setting $x'=\frac{x-x^3}{1+x^2-x^4}$, $\frac{1}{1-x'}=
552: \frac{1+x^2-x^4}{1-x+x^2+x^3-x^4}$ and
553: $\frac{x'}{1-x'}=\frac{x-x^3}{1-x+x^2+x^3-x^4}$
554: \begin{eqnarray*}
555: \sum_{n \ge 0}{\sf S}_{k,3}(n)x^n &= &
556: \frac{1}{1-x+x^2+x^3-x^4}\sum_{n \ge
557: 0}f_k(2n,0)\left(\frac{x-x^3}{1-x+x^2+x^3-x^4}\right)^{2n} \ ,
558: \end{eqnarray*}
559: whence the lemma.
560: \end{proof}
561: %%%
562: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
563: %%%
564: Using complex analysis we can extend Lemma~\ref{L:func} to arbitrary
565: $\vert z\vert < \rho_k$, where $z\in \mathbb{C}$
566: %%%
567: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
568: %%%
569: \begin{lemma}\label{L:ana}
570: Let $k>2$ be an integer, then we have for arbitrary $z\in\mathbb{C}$
571: with the property $\vert z\vert <\rho_k$ the equality
572: \begin{equation}\label{E:rr2}
573: \sum_{n \ge 0}{\sf S}_{k,3}(n)\, z^n = \frac{1}{1-z+z^2+z^3-z^4}
574: \sum_{n \ge
575: 0}f_k(2n,0)\left(\frac{z-z^3}{1-z+z^2+z^3-z^4}\right)^{2n} \ .
576: \end{equation}
577: \end{lemma}
578: %%%
579: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
580: %%%
581: \begin{proof}
582: The power series $\sum_{n\ge 0} {\sf S}_{k,3}(n) z^{n}$ and
583: $\frac{1}{1-z+z^2+z^3-z^4}
584: \sum_{n\ge 0} f_k(2n,0) \left(\frac{z-z^3}{1-z+z^2+z^3-z^4}\right)^{2n}$
585: are analytic in a disc of radius $0<\epsilon<\rho_k$ and
586: according to Lemma~\ref{L:func} coincide on the interval
587: $]-\epsilon,\epsilon [$. Therefore both functions are analytic and equal
588: on the sequence $(\frac{1}{n})_{n\in\mathbb{N}}$
589: which converges to $0$ and
590: standard results of complex analysis (zeros of nontrivial
591: analytic functions are isolated) imply that eq.~(\ref{E:rr2})
592: holds for any $z\in\mathbb{C}$ with $\vert z\vert<\rho_k$, whence
593: the lemma.
594: \end{proof}
595:
596: Lemma~\ref{L:ana} is the key to compute the exponential growth rates
597: for any $k>2$. In its proof we recruit the Theorem of Pringsheim
598: \cite{Titmarsh:39} which asserts that a power series $\sum_{n\ge
599: 0}a_nz^n$ with $a_n\ge 0$ has its radius of convergence as dominant
600: (but not necessarily unique) singularity. In particular there exists
601: a dominant real valued singularity.
602: %%%
603: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
604: %%%
605: \begin{theorem}\label{T:asy1}
606: Let $k\ge 3$ be a positive integer and $r_k$ be the radius of
607: convergence of the power series $\sum_{n\ge 0}f_k(2n,0)z^{2n}$ and
608: \begin{equation}\label{E:theta}
609: \vartheta\colon [0, \frac{\sqrt{2}}{2}]\longrightarrow
610: [0,\frac{5-\sqrt{2}}{4}], \quad z\mapsto
611: \frac{z(1-z)(1+z)}{-(z^2-\frac{1}{2})^2+z(z^2-\frac{1}{2})-
612: \frac{z}{2}+\frac{5}{4}} \ .
613: \end{equation}
614: Then the power series $\sum_{n\ge 0}{\sf S}_{k,3}(n)z^n$ has the real valued,
615: dominant singularity $\rho_k$, which is the unique real solution
616: of $\vartheta(x)=r_k$ and for the number of
617: $k$-noncrossing RNA structures with arc-length $\ge 3$ holds
618: \begin{equation}\label{E:rel}
619: {\sf S}_{k,3}(n)\sim \left(\frac{1}{\rho_k}\right)^n \ .
620: \end{equation}
621: \end{theorem}
622: %%%
623: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
624: %%%
625: In Section~\ref{S:sub} we will in particular prove that
626: $\rho_3\approx 0.21982$.
627: \begin{proof}
628: Suppose we are given $r_k$, then $r_k\le \frac{1}{2}$ (this follows
629: immediately from $C_n\sim 2^{2n}$ via Stirling's formula). The
630: functional identity of Lemma~\ref{L:func} allows us to derive the
631: radius of convergence of $\sum_{n\ge 0}S_k(n)z^n$. According to
632: Lemma~\ref{L:ana} we have
633: \begin{equation}\label{E:25}
634: \sum_{n \ge 0}{\sf S}_{k,3}(n)\, z^n = \frac{1}{1-z+z^2+z^3-z^4}
635: \sum_{n \ge
636: 0}f_k(2n,0)\left(\frac{z-z^3}{1-z+z^2+z^3-z^4}\right)^{2n}
637: \end{equation}
638: $f_k(2n,0)$ is monotone, whence the limit $\lim_{n\to
639: \infty}f_k(2n,0)^{ \frac{1}{2n}}$ exists and applying Hadamard's
640: formula: $\lim_{n\to
641: \infty}f_k(2n,0)^{\frac{1}{2n}}=\frac{1}{{r_{k}}}$. For $z\in
642: \mathbb{R}$, we proceed by computing the roots of
643: $$
644: \left|\frac{z-z^3}{1-z+z^2+z^3-z^4}\right|={r_{k}}
645: $$
646: which for $r_k\le \frac{1}{2}$ has the minimal root $\rho_k$. We
647: next show that $\rho_k$ is indeed the radius of convergence of
648: $\sum_{n\ge 0} {\sf S}_k(n) z^n$. For this purpose we observe that
649: the map
650: \begin{equation}\label{E:w=1}
651: \vartheta\colon [0, \frac{\sqrt{2}}{2}]\longrightarrow
652: [0,\frac{5-\sqrt{2}}{4}], \quad z\mapsto
653: \frac{z(1-z)(1+z)}{-(z^2-\frac{1}{2})^2+z(z^2-\frac{1}{2})-
654: \frac{z}{2}+\frac{5}{4}}
655: , \qquad \text{\rm where} \quad\vartheta(\rho_k)={r_k}
656: \end{equation}
657:
658: \begin{figure}[ht]
659: \centerline{ \epsfig{file=f7.eps,width=0.6\textwidth}\hskip15pt }
660: \caption{\small We display 4 poles(the corresponding 4 peaks) in
661: $\frac{z-z^3}{1-z+z^2+z^3-z^4}$ over $\mathbb{C}$. The picture
662: illustrates that $\vartheta$ is a bijection on the interval $[0,
663: \frac{\sqrt{2}}{2}]$, which allows us to obtain the dominant singularity
664: $\rho_k$.} \label{F:6}
665: \end{figure}
666:
667: is bijective, continuous and strictly decreasing. Continuity and
668: strict monotonicity of $\vartheta$ guarantee in view of
669: eq.~(\ref{E:w=1}) that $\rho_k$, is indeed the radius of convergence
670: of the power series $\sum_{n\ge 0} {\sf S}_k(n) z^n$. In order to
671: show that $\rho_k$ is a dominant singularity we consider $\sum_{n\ge
672: 0}{\sf S}_{k,3}(n)z^n$ as a power series over $\mathbb{C}$. Since
673: ${\sf S}_{k,3}(n)\ge 0$, the theorem of
674: Pringsheim~\cite{Titmarsh:39} guarantees that $\rho_k$ itself is a
675: singularity. By construction $\rho_k$ has minimal absolute value and
676: is accordingly dominant. Since ${\sf S}_{k,3}(n)$ is monotone
677: $\lim_{n\to\infty}{\sf S}_{k,3}(n)^{\frac{1} {n}}$ exists and we
678: obtain using Hadamard's formula
679: \begin{equation}
680: \lim_{n\to\infty}{\sf
681: S}_{k,3}(n)^{\frac{1}{n}}=\frac{1}{\rho_k},\quad \text{\rm or
682: equivalently}\quad {\sf S}_{k,3}(n)\sim \left(\frac{1}{\rho_k}
683: \right)^n \, ,
684: \end{equation}
685: from which eq.~(\ref{E:rel}) follows and the proof of the theorem is complete.
686: \end{proof}
687: %%%%
688: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
689: %%%
690: \section{Asymptotic Analysis}\label{S:sub}
691: %%%
692: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
693: %%%
694:
695: In this section we provide the asymptotic number of $3$-noncrossing RNA
696: structures with arc-length $\ge 3$.
697: In the course of our analysis we derive the analytic continuation of the
698: power series $\sum_{n\ge 0}{\sf S}_{3,3}(n)z^n$. The analysis will in
699: particular provide independent proof of the exponential factor computed
700: in Theorem~\ref{T:asy1}. The derivation of the subexponential factors is
701: based on singular expansions \cite{Flajolet:05} in combination with transfer
702: theorems. The key ingredient for the coefficient extraction is the Hankel
703: contours, see Figure~\ref{F:7}. Let us begin by specifying a
704: suitable domain for our Hankel contours tailored for
705: Theorem~\ref{T:transfer1}.
706:
707: \begin{definition}\label{D:delta}
708: Given two numbers $\phi,R$, where $R>1$ and $0<\phi<\frac{\pi}{2}$
709: and $\rho\in\mathbb{R}$ the open domain $\Delta_\rho(\phi,R)$ is
710: defined as
711: \begin{equation}
712: \Delta_\rho(\phi,R)=\{ z\mid \vert z\vert < R, z\neq \rho,\, \vert
713: {\rm Arg}(z-\rho)\vert >\phi\}
714: \end{equation}
715: A domain is a $\Delta_\rho$-domain if it is of the form
716: $\Delta_\rho(\phi,R)$ for some $R$ and $\phi$. A function is
717: $\Delta_\rho$-analytic if it is analytic in some
718: $\Delta_\rho$-domain.
719: \end{definition}
720: %%%
721: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
722: %%%
723: %%%%
724: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
725: %%%%
726: \begin{figure}[ht]
727: \centerline{%
728: \epsfig{file=f8.eps,width=0.5\textwidth}\hskip15pt } \caption{\small
729: $\Delta_1$-domain enclosing a Hankel contour. We assume $z=1$ to be
730: the unique dominant singularity. The coefficients are obtained via
731: Cauchy's integral formula and the integral path is decomposed in $4$
732: segments. Segment $1$ becomes asymptotically irrelevant since by
733: construction the function involved is bounded on this segment.
734: Relevant are the rectilinear segments $2$ and $4$ and the inner
735: circle $3$. The only contributions to the contour integral are being
736: made here, which shows why the singular expansion allows to
737: approximate the coefficients so well.} \label{F:7}
738: \end{figure}
739: %%%%
740: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
741: %%%%
742: Since the Taylor coefficients have the property
743: \begin{equation}\label{E:scaling}
744: \forall \,\gamma\in\mathbb{C}\setminus 0;\quad [z^n]f(z)=\gamma^n
745: [z^n]f(\frac{z}{\gamma}) \ ,
746: \end{equation}
747: we can, w.l.o.g.~reduce our analysis to the case where $1$ is the
748: dominant singularity. We use $U(a,r)=\{z\in \mathbb{C}|\vert
749: z-a\vert<r\}$ to denote the open neighborhood of $a$ in
750: $\mathbb{C}$.
751:
752:
753: We use the notation
754: \begin{equation}\label{E:genau}
755: \left(f(z)=O\left(g(z)\right) \ \text{\rm as $z\rightarrow
756: \rho$}\right)\quad \Longleftrightarrow \quad \left(f(z)/g(z) \
757: \text{\rm is bounded as $z\rightarrow \rho$}\right)
758: \end{equation}
759: and if we write $f(z)=O(g(z))$ it is implicitly assumed that $z$
760: tends to a (unique) singularity. $[z^n]\,f(z)$ denotes the
761: coefficient of $z^n$ in the power series expansion of $f(z)$ around
762: $0$.
763: %%%
764: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
765: %%%
766: \begin{theorem}\label{T:transfer1}\cite{Flajolet:05}
767: Let $\alpha$ be an arbitrary complex number in $\mathbb{C}\setminus
768: \mathbb{Z}_{\le 0}$, $f(z)$ be a $\Delta_1$-analytic function.
769: Suppose $r\in\mathbb{Z}_{\ge 0}$, and
770: $f(z)=O((1-z)^{r}\ln^{}(\frac{1}{1-z}))$ in the intersection of a
771: neighborhood of $1$ and the $\Delta_1$-domain, then we have
772: \begin{equation}
773: [z^n]f(z)\sim K\, (-1)^r\frac{r!}{n(n-1)\dots(n-r)} \quad \text{\it
774: for some $K>0$}\ .
775: \end{equation}
776: \end{theorem}
777: %%%
778: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
779: %%%
780:
781: We are now prepared to compute an explicit formula for the numbers
782: of $3$-noncrossing RNA structures with arc-length $\ge 3$.
783:
784: %%%
785: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
786: %%%
787: \begin{theorem}\label{T:asy3}
788: The number of $3$-noncrossing RNA structures with arc length $\ge 3$
789: is asymptotically given by
790: \begin{eqnarray*}
791: \label{E:konk3} {\sf S}_{3,3}(n) & \sim & \frac{6.11170\cdot
792: 4!}{n(n-1)\dots(n-4)}\,
793: 4.54920^n \ .\\
794: \end{eqnarray*}
795: \end{theorem}
796: %%%
797: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
798: %%%
799: \begin{proof}
800: {\it Claim $1$.}
801: The dominant singularity $\rho_3$ of the power series
802: $\sum_{n\ge 0} {\sf S}_{3,3}(n) z^n$ is unique.\\
803: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
804: In order to prove Claim $1$ we use Lemma~\ref{L:ana}, according to
805: which the analytic function $\Xi_3(z)$ is the analytic continuation
806: of the power series $\sum_{n\ge 0} {\sf S}_{3}(n) z^n$. We proceed
807: by showing that $\Xi_3(z)$ has exactly $12$ singularities in
808: $\mathbb{C}$ and the dominant singularity is unique. The first four
809: singularities are the roots of the quartic polynomial
810: $P(z)=1-z+z^2+z^3-z^4$. Next we observe in analogy to our proof
811: in \cite{Reidys:07asy} that the power series
812: $\sum_{n\ge 0} f_3(2n,0) y^{n}$ has the analytic continuation
813: $\Psi(y)$ (obtained by MAPLE sumtools) given by
814: \begin{equation}\label{E:psi}
815: \Psi(y)=
816: \frac{-(1-16y)^{\frac{3}{2}} P_{3/2}^{-1}(-\frac{16y+1}{16y-1})}
817: {16\, {y}^{\frac{5}{2}}} \ ,
818: \end{equation}
819: where $P_{\nu}^{m}(x)$ denotes the Legendre Polynomial of the first
820: kind with the parameters $\nu=\frac{3}{2}$ and $m=-1$. $\Psi(y)$ has
821: one dominant singularity at $y=\frac{1}{16}$, which in view of
822: $\vartheta(z)=(\frac{z-z^3}{1+z^2-z^4-z+z^3})^2$ induces exactly $8$
823: singularities of
824: $\Xi_3(z)=\frac{1}{1+z^2-z^4-z+z^3}\,
825: \Psi\left(\left(\frac{z-z^3}{1+z^2-z^4-z+z^3}\right)^2\right)
826: $. Indeed, $\Psi(y^2)$ has the two singularities $\mathbb{C}$:
827: $\beta_1=\frac{1}{4}$ and $\beta_2=-\frac{1}{4}$ which produces
828: for $\Xi_3(z)$ (solving the quartic equation) the 8 singularities
829: $\rho_3\approx 0.21982$, $\zeta_2\approx
830: 5.00829$, $\zeta_3\approx -1.07392$, $\zeta_4\approx 0.84581$,
831: $\zeta_5\approx -0.53243+0.11951i$, $\zeta_6 \approx -0.53243-0.11951i$,
832: $\zeta_7 \approx 1.10477$ and $\zeta_8 \approx -3.03992$.
833: The above values have error terms
834: (for details for solving the general quartic equation see Section~\ref{S:app})
835: of the order $10^{-5}$, which allows us
836: to conclude that the dominant singularity $\rho_3$ is unique and Claim
837: $1$ follows. \\
838: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
839: {\it Claim $2$.} \cite{Reidys:07asy} $\Psi(z)$ is
840: $\Delta_{\frac{1}{16}}(\phi,R)$-analytic and has the singular
841: expansion $(1-16z)^4\ln\left(\frac{1}{1-16z}\right)$.
842: \begin{equation}
843: \forall\, z\in\Delta_{\frac{1}{16}}(\phi,R)\cap
844: U(\frac{1}{16},\epsilon);\quad
845: \Psi(z)={O}\left((1-16z)^4\ln\left(\frac{1}{1-16z}\right)\right) \ .
846: \end{equation}
847: First $\Delta_{\frac{1}{16}}(\phi,R)$-analyticity of the function
848: $\Psi(z)$ is obvious.
849: We proceed by proving that $(1-16z)^4\ln\left(\frac{1}{1-16z}\right)$ is
850: its singular expansion in the intersection of a neighborhood of $\frac{1}{16}$
851: and the $\Delta$-domain $\Delta_{\frac{1}{16}}(\phi,R)$.
852: Using the notation of falling factorials
853: $(n-1)_4=(n-1)(n-2)(n-3)(n-4)$ we observe
854: $$
855: f_3(2n,0)=C_{n+2}C_{n}-C_{n+1}^2= \frac{1}{(n-1)_4}
856: \frac{12(n-1)_4(2n+1)}{(n+3)(n+1)^2(n+2)^2}\,
857: \binom{2n}{n}^2 \ .
858: $$
859: With this expression for $f_3(2n,0)$ we arrive at the formal identity
860: \begin{eqnarray*}
861: \sum_{n\ge 5}16^{-n}f_3(2n,0)z^n & = &
862: O(\sum_{n\ge 5}
863: \left[16^{-n}\,\frac{1}{(n-1)_4}
864: \frac{12(n-1)_4(2n+1)}{(n+3)(n+1)^2(n+2)^2}\,
865: \binom{2n}{n}^2-\frac{4!}{(n-1)_4}\frac{1}{\pi}\frac{1}{n}\right]z^n \\
866: & & + \sum_{n\ge 5}\frac{4!}{(n-1)_4}\frac{1}{\pi}\frac{1}{n}z^n) \ ,
867: \end{eqnarray*}
868: where $f(z)=O(g(z))$ denotes that the limit $f(z)/g(z)$ is bounded
869: for $z\rightarrow 1$, eq.~(\ref{E:genau}). It is clear that
870: \begin{eqnarray*}
871: & & \lim_{z\to 1}(\sum_{n\ge 5}\left[16^{-n}\,\frac{1}{(n-1)_4}
872: \frac{12(n-1)_4(2n+1)}{(n+3)(n+1)^2(n+2)^2}\,
873: \binom{2n}{n}^2-\frac{4!}{(n-1)_4}\frac{1}{\pi}\frac{1}{n}\right]z^n) \\
874: &= &
875: \sum_{n\ge 5} \left[16^{-n}\,\frac{1}{(n-1)_4}
876: \frac{12(n-1)_4(2n+1)}{(n+3)(n+1)^2(n+2)^2}\,
877: \binom{2n}{n}^2-\frac{4!}{(n-1)_4}\frac{1}{\pi}\frac{1}{n}\right]
878: <\kappa
879: \end{eqnarray*}
880: for some $\kappa< 0.0784$. Therefore we can conclude
881: \begin{equation}
882: \sum_{n\ge 5}16^{-n}f_3(2n,0)z^n=
883: O(\sum_{n\ge 5}\frac{4!}{(n-1)_4}\frac{1}{\pi}\frac{1}{n}z^n) \ .
884: \end{equation}
885: We proceed by interpreting the power series on the rhs, observing
886: \begin{equation}
887: \forall\, n\ge 5\, ; \qquad
888: [z^n]\left((1-z)^4\,\ln\frac{1}{1-z}\right)=
889: \frac{4!}{(n-1)\dots (n-4)}\frac{1}{n} \, ,
890: \end{equation}
891: whence $\left((1-z)^4\,\ln\frac{1}{1-z}\right)$ is
892: the unique analytic continuation of $\sum_{n\ge 5}\frac{4!}{(n-1)_4}
893: \frac{1}{\pi}\frac{1}{n}z^n$.
894: Using the scaling property of Taylor coefficients
895: $[z^n]f(z)=\gamma^n [z^n]f(\frac{z}{\gamma})$ we obtain
896: \begin{equation}\label{E:isses}
897: \forall\, z\in\Delta_{\frac{1}{16}}(\phi,R)\cap
898: U(\frac{1}{16},\epsilon);\quad \Psi(z)
899: =O\left((1-16z)^4\ln\left(\frac{1}{1-16z}\right)\right) \ .
900: \end{equation}
901: Therefore we have proved that $(1-16z)^{4}\ln^{}(\frac{1}{1-16z})$
902: is the singular expansion of $\Psi(z)$ at $z=\frac{1}{16}$, whence
903: Claim $2$. Our last step consists in verifying that the type of the
904: singularity does not change when passing from $\Psi(z)$ to
905: $\Xi_3(z)= \frac{1}{1-z+z^2+z^3-z^4}
906: \Psi((\frac{z-z^3}{1-z+z^2+z^3-z^4})^2)$.
907: \\
908: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
909: {\it Claim $3$.} For $z\in
910: \Delta_{\rho_3}(\phi,R)\cap U(\rho_3,\epsilon)$ we have $\Xi_3(z)
911: ={O}\left((1-\frac{z}{\rho_3})^4\ln(\frac{1}{1-\frac{z}{\rho_3}})
912: \right)$.\\
913: To prove the claim set $u(z)=1-z+z^2+z^3-z^4$. We first observe that
914: Claim $2$ and Lemma~\ref{L:ana} imply
915: \begin{align*}
916: \Xi_3(z) &=O\left(
917: \frac{1}{u(z)}\,
918: \left[\left(1-16(\frac{z-z^3}{u(z)})^2\right)^4
919: \ln\frac{1}{\left(1-16(\frac{z-z^3}{u(z)})^2\right)}\right]\right)
920: \ .
921: \end{align*}
922: The Taylor expansion of $q(z)=1-16(\frac{z-z^3}{u(z)})^2$ at
923: $\rho_3$ is given by $q(z)=\alpha(\rho_3-z)+{O}(z-\rho_3)^2$ and
924: setting $\alpha\approx -1.15861$ we compute
925: \begin{align*}
926: \frac{1}{u(z)}\, \left[q(z)^4\ln\frac{1} {q(z)}\right] &=
927: \frac{(\alpha(\rho_3-z)+{O}(z-\rho_3)^2)^4\ln\frac{1}{\alpha(\rho_3-z)+{O}
928: (z-\rho_3)^2}}{0.83679-0.45789(z-\rho_3)+O((z-\rho_3)^2)}\\
929: &= \frac{\left([\alpha+O(z-\rho_3)](\rho_3-z)^4
930: \ln\frac{1}{[\alpha+O(z-\rho_3)](\rho_3-z)} \right)}{O(z-\rho_3)}
931: \\
932: &={O}((\rho_3-z)^4\ln\frac{1}{\rho_3-z}) \ ,
933: \end{align*}
934: whence Claim $3$. Now we are in the position to employ
935: Theorem~\ref{T:transfer1}, and obtain for ${\sf S}_{3,3}(n)$
936: \begin{align*}
937: {\sf S}_{3,3}(n)&\sim K'\,
938: [z^n]\left((\rho_3-z)^4\ln\frac{1}{\rho_3-z} \right) \sim K'\,
939: \frac{4!} {n(n-1)\dots(n-4)}\left(\frac{1}{\rho_3}\right)^n \ .
940: \end{align*}
941: Theorem~\ref{T:cool2} allows us to compute $K'=6.11170$ and the proof of
942: the Theorem is complete.
943: \end{proof}
944: %%%
945: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
946: %%%
947:
948: \section{Appendix}\label{S:app}
949:
950: %%%
951: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
952: %%%
953:
954: Let us first introduce some basic definitions for quartic equation
955: (quartic) used in the following. $Ax^4+Bx^3+Cx^2+Dx+E=0$ is
956: a quartic if $A \ne 0$. A depressed quartic is a quartic such that
957: $B=0$ holds. $Ax^3+Bx^2+Cx+d=0$ is a cubic equation if $A \ne 0$ and in
958: particular a cubic without $x^2$ term is called as depressed cubic.
959: \\
960: The first step in solving the quartic consists in transforming it into
961: a depressed quartic. i.e. eliminate the $x^3$ term. Transform the
962: equation into $x^4+\frac{B}{A}x^3+\frac{C}{A}x^2+\frac{D}{A}x+\frac{E}{A}=0$
963: and substitute $x=u-\frac{B}{4A}$. Simplifying the original quartic yields
964: \begin{equation}\label{A:depqua}
965: u^4+\alpha u^2+\beta u +\gamma=0
966: \end{equation}
967: where $\alpha=\frac{-3B^2}{8A^2}+\frac{C}{A}$,
968: $\beta=\frac{B^3}{8A^3}-\frac{BC}{2A^2}+\frac{D}{A}$ and
969: $\gamma=\frac{-3B^4}{256A^4}+\frac{CB^2}{16A^3}-\frac{BD}{4A^2}
970: +\frac{E}{A}$, and eq.~(\ref{A:depqua}) is a depressed quartic
971: function, which is tantamount to
972: \begin{equation}\label{A:preu}
973: (u^2+\alpha)^2+\beta u+\gamma=\alpha u^2+\alpha^2
974: \end{equation}
975: The next step is to insert a variable $y$ into the perfect square
976: on the left side of eq.~(\ref{A:preu}). Add both
977: \begin{align*}
978: (u^2+\alpha+y)^2-(u^2+\alpha)^2=2yu^2+2y\alpha+y^2\\
979: 0=(\alpha+2y)u^2-2yu^2-\alpha u^2
980: \end{align*}
981: to eq.(\ref{A:preu}) yields
982: \begin{equation}\label{A:insert-y}
983: (u^2+\alpha+y)^2=(\alpha+2y)u^2-\beta
984: u+(y^2+2y\alpha+\alpha^2-\gamma) \ .
985: \end{equation}
986: The next step consists in substituting for $y$ such that the right
987: side of eq.~(\ref{A:insert-y}) becomes a square. Observe that
988: $(su+t)^2=(s^2)u^2+(2st)u+(t^2)$ holds for any $s$ and $t$, and
989: the relation between the coefficients of the rhs is
990: $(2st)^2=4(s^2)(t^2)$. Therefore to make the rhs of
991: eq.~(\ref{A:insert-y}) into a perfect square, the following
992: equation must hold.
993: \begin{equation}{\label{A:solve-y}}
994: 2y^3+5 \alpha y^2+(4 \alpha^2-2 \gamma)y+(\alpha^3-\alpha
995: \gamma-\frac{\beta^2}{4})=0 \ .
996: \end{equation}
997: Similarly, transform eq~(\ref{A:solve-y}) into a
998: depressed cubic equation by substituting $y=v-\frac{5}{6}\alpha$
999: \begin{equation}\label{A:depcub}
1000: v^3+(-\frac{\alpha^2}{12}-\gamma)v+(-\frac{\alpha^3}{108}+\frac{\alpha
1001: \gamma}{3}-\frac{\beta^2}{8})=0 \ .
1002: \end{equation}
1003: Set $P=-\frac{\alpha^2}{12}-\gamma$
1004: and $Q=-\frac{\alpha^3}{108}+\frac{\alpha
1005: \gamma}{3}-\frac{\beta^2}{8}$. Select any of the solutions of
1006: eq.~(\ref{A:depcub}) of the form $v=\frac{P}{3U}-U$ where
1007: $U=\sqrt[3]{\frac{Q}{2}\pm\sqrt{\frac{Q^2}{4}+\frac{P^3}{27}}}$
1008: and $U \ne 0$, otherwise $v=0$. In view of $y=v-\frac{5}{6}$ this solution
1009: yields for eq.~(\ref{A:solve-y})
1010: $y=-\frac{5}{6}\alpha+\frac{P}{3U}-U$ for $U \ne 0$ and
1011: $y=-\frac{5}{6}\alpha$ for $U=0$. Now the rhs of
1012: eq.~(\ref{A:insert-y}) becomes
1013: \begin{align*}
1014: (\alpha+2y)u^2+(-\beta)u+(y^2+2y\alpha+\alpha^2-\gamma)=\left(
1015: \left(\sqrt{\alpha+2y}\right)u+\frac{-\beta}{2\sqrt{\alpha+2y}}\right)^2
1016: \end{align*}
1017: Combined with eq.~(\ref{A:insert-y}) this yields two solutions for $u$:
1018: $$
1019: u=\frac{\pm\sqrt{\alpha+2y}\pm\sqrt{-\left(3\alpha+2y\pm\frac{2\beta}
1020: {\sqrt{\alpha+2y}}\right)}}{2}
1021: $$
1022: where the first and the third $\pm$ must have the same sign.
1023: This allows us to obtain the solutions for $x$:
1024: $$
1025: x=-\frac{B}{4A}+\frac{\pm\sqrt{\alpha+2y}\pm\sqrt{-\left(3\alpha+
1026: 2y\pm\frac{2\beta}{\sqrt{\alpha+2y}}\right)}}{2} \ .
1027: $$
1028: In particular for $x^4-5x^3-x^2+5x-1=0$, $A=1,B=-5,C=-1,D=5,E=-1$,
1029: and hence $\alpha=-\frac{83}{8}$, $\beta=-\frac{105}{8}$,
1030: $P=-\frac{16}{3}$ and $Q=\frac{299}{216}$ and $U \approx
1031: 1.21481-0.54955i \ne 0$, therefore $y \approx 6.21621$, and the
1032: solutions are $\rho_3\approx 0.21982$, $\zeta_2\approx 5.00829$,
1033: $\zeta_3\approx -1.07392$, $\zeta_4\approx 0.84581$. As for the
1034: equation $x^4+3x^3-x^2-3x-1=0$, the corresponding solutions are
1035: $\zeta_5\approx -0.53243+0.11951i$, $\zeta_6 \approx
1036: -0.53243-0.11951i$, $\zeta_7 \approx 1.10477$ and $\zeta_8 \approx
1037: -3.03992$.
1038:
1039:
1040: %%%
1041: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1042: %%%
1043: {\bf Acknowledgments.}
1044: %%%
1045: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1046: %%%
1047: %We are grateful to ??? for helpful comments.
1048: This work was supported by the
1049: 973 Project, the PCSIRT Project of the Ministry of Education, the Ministry
1050: of Science and Technology, and the National Science Foundation of China.
1051:
1052:
1053: \bibliography{a2}
1054: \bibliographystyle{plain}
1055:
1056:
1057: %%%
1058: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1059: %%%
1060:
1061: \end{document}
1062: