ViennaRNA 安装使用

RNAfold 是 Vienna RNA Package 里的一种通过计算最小自由能 (MFE: minimum free energy) 来预测 RNA 二级结构的软件。它的算法和理论支持主要来源于 M. Zuker 和 P. Stiegler 在 1981 年发表的这篇 Optimal computer folding of large RNA sequences using thermodynamic and auxiliary information。想了解 Michael Zuker 教授小组的一系列研究情况和 MFE 算法计算依据的可以点这里。

主要参数¶

RNAfold -p -d2 --noLP < sequence1.fa > sequence1.out

详细的参数参考：https://www.tbi.univie.ac.at/RNA/RNAfold.1.html。本节部分内容来源于：RNAfold 介绍 - 360doc

-p, –partfunc[=INT] Calculate the partition function and base pairing probability matrix. (default=‘1’)
-noLP 表示不允许有孤立碱基对；
-noPS 表示不生成结构图（只计算结果，即不产生二级结构对应的 postscript 文件，这种文件可以转换为 PDF 格式）；
d1 指的是 “dangling end” (悬挂端)的能量计算按照 d1 模式（有且仅有未配对的碱基可以参与最多一个悬挂端），返回值就是“点-弧”结构图和预测的能量值。

关于 dangling end(dangling bases) 的中文翻译在国内未能查到，这里暂翻译为“悬挂端”。可以参考这篇 Long RNA Dangling End Has Large Energetic Contribution to Duplex Stability，dangling end 指的就是在双链中未配对、会参与旁边“链栈”(stacks) 能量增益的碱基。不同的悬挂组合产生的不同的能量增益，这在 RNAfold 计算中也会被考虑（寻求最好的 dangle）。

输出结果¶

默认参数会输出以下两种二级结构：

optimal secondary structure。最佳二级结构，保证对应的自由能最小，最小自由能简称 MFE，结果示意如下
centroid secondary structure。自由能表征改变这个结构需要注入的能量大小，对应的数值越小，该结构越稳定。

安装¶

ViennaRNA-2.5.x 要求 C++17 标准的更高版本的编译器，即 GCC 需要 5 及以上版本。

That is unfortunate! We added dlib to our source tree for version 2.5.0 to efficiently compute concentration dependencies upon multi-strand interactions. Apparently, dlib requires a more recent compiler that actually implements the C++17 standard, which is only available since GCC 5. —— From issues 146 我们在版本 2.5.0 中将 dlib 添加到了我们的源代码树中，以便高效地计算多股链交互的浓度依赖性。显然，dlib 需要一个实现了 C++17 标准的更高版本的编译器，这仅在 GCC 5 之后可用。

$ ./configure --prefix=/Bioinfo/Pipeline/SoftWare/ViennaRNA-2.4.11
......
configure:

======================================
    ViennaRNA Package 2.4.11
======================================

Sub Packages
------------
  * Kinfold                   : yes
  * RNAforester               : yes
  * Analyse{Dists,Seqs}       : no
  * RNAlocmin                 : yes
  * Kinwalker                 : no

Extra Libraries
---------------
  * Support Vector Machine    : yes
  * GNU Scientific Library    : no
  * GNU MPFR                  : no
  * JSON                      : yes

Features
--------
  * Boustrophedon             : yes
  * Use hash for NR Sampling  : no
  * C11 features              : yes
  * TTY colors                : yes
  * Float Precision(PF}       : no
  * Deprecation Warnings      : no

Optimizations
-------------
  * Auto Vectorization        : yes
  * Explicit SIMD Extension   : yes
  * Link Time Optimization    : yes
  * POSIX Threads             : yes
  * OpenMP                    : yes

Scripting Language Interfaces
-----------------------------
  * Perl 5                    : yes
  * Python 2                  : yes
  * Python 3                  : yes

Documentation
-------------
  * Reference Manual (PDF)    : yes
  * Reference Manual (HTML)   : yes
  * Tutorial (PDF)            : yes
  * Tutorial (HTML)           : no

Unit Tests
----------
  * Executable Programs       : yes
  * C-Library                 : no
  * Perl 5 Interface          : yes
  * Python 2 Interface        : yes
  * Python 3 Interface        : yes

MacOS X
-------
  * Universal Binary          : no
  * Installer                 : no
  * SDK                       : custom

Install Directories
-------------------
  * Executables               : /galaxy-dist/software/ViennaRNA-2.4.11/bin
  * Libraries                 : /galaxy-dist/software/ViennaRNA-2.4.11/lib
  * Header files              : /galaxy-dist/software/ViennaRNA-2.4.11/include
  * Extra Data                : /galaxy-dist/software/ViennaRNA-2.4.11/share
  * Man pages                 : /galaxy-dist/software/ViennaRNA-2.4.11/share/man
  * Documentation             : /galaxy-dist/software/ViennaRNA-2.4.11/share/doc/ViennaRNA
      (HTML)                  : /galaxy-dist/software/ViennaRNA-2.4.11/share/doc/ViennaRNA/html
      (PDF)                   : /galaxy-dist/software/ViennaRNA-2.4.11/share/doc/ViennaRNA
  * Perl5 Interface           :
      (binaries)              : /galaxy-dist/software/ViennaRNA-2.4.11/lib/perl5/site_perl/5.26.1/x86_64-linux-thread-multi
      (scripts)               : /galaxy-dist/software/ViennaRNA-2.4.11/lib/perl5/site_perl/5.26.1
  * Python2 Interface         :
      (binaries)              : /galaxy-dist/software/ViennaRNA-2.4.11/lib/python2.7/site-packages
      (scripts)               : /galaxy-dist/software/ViennaRNA-2.4.11/lib/python2.7/site-packages
  * Python3 Interface         :
      (binaries)              : /galaxy-dist/software/ViennaRNA-2.4.11/lib/python3.6/site-packages
      (scripts)               : /galaxy-dist/software/ViennaRNA-2.4.11/lib/python3.6/site-packages

You can run 'make', 'make check' and 'make install' now!
$ make
$ make install

命令行使用示例。

RNAfold -p -d2 --noLP < sequence1.fa > sequence1.out

使用示例¶

RNAlib Python¶

参考教程：

import RNA
seq = "GGGCUAUUAGCUCAGUUGGUUAGAGCGCACCCCUGAUAAGGGUGAGGUCGCUGAUUCGAAUUCAGCAUAGCCCA"
md = RNA.md()

# change temperature and dangle model
md.temperature = 20.0 # 20 Deg Celcius
md.dangles     = 1    # Dangle Model 1
md.noLP              = 1

fc = RNA.fold_compound(seq, md)
(ss, mfe) = fc.mfe()
print("{}\n{} ({:6.2f})".format(seq, ss, mfe))

GGGCUAUUAGCUCAGUUGGUUAGAGCGCACCCCUGAUAAGGGUGAGGUCGCUGAUUCGAAUUCAGCAUAGCCCA
(((((((..((((.........)))).(((((.......))))).....(((((.......)))))))))))). (-28.90)

RNAfold web server¶

在线地址：http://rna.tbi.univie.ac.at//cgi-bin/RNAWebSuite/RNAfold.cgi

示例 1：

# Results have been computed using RNAfold 2.4.17. An equivalent command line call would have been
RNAfold -p -d0 --noClosingGU < test_sequenc.fa > test_sequenc.out

示例 2：

# Results have been computed using RNAfold 2.4.17. An equivalent command line call would have been
RNAfold -p -d0 --noLP --noClosingGU < test_sequenc.fa > test_sequenc.out

示例 3：

# Results have been computed using RNAfold 2.4.17. An equivalent command line call would have been
RNAfold -p -d0 --noLP < test_sequenc.fa > test_sequenc.out

示例 4：

# Results have been computed using RNAfold 2.4.17. An equivalent command line call would have been
RNAfold -d0 --noLP --noClosingGU < test_sequenc.fa > test_sequenc.out