1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
|
<!DOCTYPE book [
<!ENTITY % sgml.features "IGNORE">
<!ENTITY % xml.features "INCLUDE">
<!ENTITY % ent-mmlalias
PUBLIC "-//W3C//ENTITIES Aiases for MathML 2.0//EN"
"/usr/share/xml/schema/w3c/mathml/dtd/mmlalias.ent" >
%ent-mmlalias;
]>
<article xmlns="http://docbook.org/ns/docbook" version="5.0"
xmlns:mml="http://www.w3.org/1998/Math/MathML"
xml:lang="en">
<info><title>Alignment Practical Exercises</title>
<author><personname><firstname>Benjamin</firstname><surname>Redelings</surname></personname></author>
</info>
<para>This tutorial assumes a UNIX command line. On Linux or Mac this is built in; on Windows install Cygwin.</para>
<section><info><title>Exercise (Setup, MAFFT, HoT, AliView)</title></info>
<section><info><title>Make a <filename>~/bin</filename> directory (for programs) and add it to your PATH</title></info>
% cd // go to your home directory
% pwd // where are we? = print working directory
% ls // look around
% mkdir ~/bin // make a subdirectory
% ls .bash_profile .profile // do you have .profile or .bash_profile?
% nano .profile // edit the startup script
%% OR
% nano .bash_profile // edit the startup script
%%<Add the line "PATH=~/bin:$PATH".>
% cat ~/.bash_profile // Check if your changes are there
% cat ~/.profile // Check if your changes are there
%%<Log out and log back in.>
% echo $PATH // check that ~/bin is in PATH
% echo $PATH | sed 's/:/\n/g' // easier to read!
</section>
<section><info><title>Download the alignment files we will use</title></info>
% cd
% mkdir alignment_files
% cd ~/alignment_files
% wget http://www.bali-phy.org/examples.tgz // Download the files - Linux
% curl -O http://www.bali-phy.org/examples.tgz // Download the files - Mac
%%<Alternatively, download the files using your browser>
% tar -zxf examples.tgz // extract the compressed archive
% ls examples // look inside the new directory that was created
% less examples/ferns/cleaned.fasta // examine the downloaded files – type q to quit
% alignment-info examples/ferns/cleaned.fasta
</section>
<section><info><title>Install fasta-flip.pl to ~/bin</title></info>
% cd ~/alignment_files/examples
% chmod +x fasta-flip.pl // make the file executable
% cp fasta-flip.pl ~/bin // copy the file to ~/bin
% which fasta-flip.pl // check to make sure its in your PATH
</section>
<section><info><title>MAFFT and HoT and Seaview</title></info>
% cp ferns/cleaned.fasta ferns.fasta // make a copy of a file
% mafft --auto ferns.fasta > ferns-mafft.fasta // try a simple sequence alignment
% fasta-flip.pl ferns.fasta | mafft --auto - | fasta-flip.pl > ferns-mafft-flipped.fasta
% aliview ferns-mafft.fasta & aliview ferns-mafft-flipped.fasta & // take a look at the two good alignments
% ./compare-alignments ferns-mafft.fasta ferns-mafft-flipped.fasta > ferns-compare.html
% firefox ferns-compare.html // load in firefox, then shrink the font
</section>
</section>
<section><info><title>Exercise (Iterative refinement, FSA, PRANK)</title></info>
<section><info><title>Refinement in MAFFT</title></info>
Look at the time to perform different numbers of iterations with MAFFT:
% cd ~/alignment_files/examples
% time mafft --localpair --maxiterate 0 ferns.fasta > ferns-mafft-0.fasta // No refinement
% time mafft --localpair --maxiterate 1 ferns.fasta > ferns-mafft-1.fasta // Refine 1 times
% time mafft --localpair --maxiterate 1000 ferns.fasta > ferns-mafft-2.fasta // Refine until done
% aliview ferns-mafft-0.fasta & aliview ferns-mafft-1.fasta & // take a look at the two alignments
% ./compare-alignments ferns-mafft-0.fasta ferns-mafft-1.fasta > ferns-compare-0-1.html
% firefox ferns-compare-0-1.html
</section>
<!-- Issue the ferns introns really are unalignable! -->
<section><info><title>Alignment with FSA</title></info>
Lets run FSA with no penalty for wrong homologies, like ProbCons.
% time fsa --fast --gapfactor 0 ferns.fasta > ferns-fsa0.fasta // This is similar to probcons – few gaps.
% alignment-info ferns-fsa0.fasta | grep columns
% aliview ferns-fsa0.fasta &
Let's run FSA with the standard penalty for wrong homologies:
% time fsa --fast ferns.fasta > ferns-fsa.fasta // This is the default. Some gaps.
% alignment-info ferns-fsa.fasta | grep columns
% aliview ferns-fsa.fasta &
How long is the probcons-like alignment, versus the FSA alignment?
</section>
<section><info><title>Alignment with PRANK</title></info>
Now let's run PRANK:
% time prank -d=ferns.fasta +F -o=ferns-prank
% cp ferns-prank.best.fas ferns-prank.fasta
% alignment-info ferns-prank.fasta | grep columns
% aliview ferns-prank.fasta &
How long is the PRANK alignment? How is the shape different from the other alignments?
</section>
</section>
<section><info><title>BAli-Phy</title></info>
<para>This exercise offers a quick overview of bali-phy with an emphasis on concrete analysis of real datasets. Other documentation includes:
<itemizedlist>
<listitem>The <link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="http://www.bali-phy.org/README.xhtml">Users Guide</link> is much more complete.</listitem>
<listitem>The <link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="http://www.bali-phy.org/Tutorial4.xhtml">Tutorial</link> is similar to this exercise, but covers more functionality.</listitem>
<listitem>You can run <userinput><replaceable>tool</replaceable> --help</userinput> to see command line options for <replaceable>tool</replaceable>.</listitem>
<listitem>You can run <userinput>man <replaceable>tool</replaceable></userinput> to see the manual page for <replaceable>tool</replaceable>.</listitem>
<listitem>Manual pages for bali-phy and tools are also available <link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="http://www.bali-phy.org/docs.php#manpages">online</link>.</listitem>
</itemizedlist>
</para>
<section><info><title>Creating a script file</title></info>
<para>First, go to the <filename>~/alignment_files/examples</filename> directory and create
a text file called ferns-intron-exon.txt that looks like this:</para>
<programlisting># sequence data for 7 partitions
align = ferns.fasta:5-8 # exon 1
align = ferns.fasta:9-453 # intron 1
align = ferns.fasta:454-596 # exon 2
align = ferns.fasta:597-864 # intron 2
align = ferns.fasta:865-948 # exon 3
align = ferns.fasta:949-1328 # intron 3
align = ferns.fasta:1329-1393 # exon 4
smodel = gtr+Rates.gamma+inv # substitution model
name = ferns-intron-exon # determines name for output directory
</programlisting>
<para>It should be possible to copy and paste this text into the nano editor below
% cd ~/alignment_files/examples
% nano ferns-intron-exon.txt // Note the instructions on the bottom of the terminal: <userinput>^X</userinput> means <userinput>Control-X</userinput>
% cat ferns-intron-exon.txt
% bali-phy -c ferns-intron-exon.txt --test | less
</para>
</section>
<section><info><title>Command line options</title></info>
<para>A script file can be turned into a command-line by replacing commands like <userinput>command = value</userinput> with <userinput>--command value</userinput>. The exception is that for <userinput>align=file</userinput> commands, you can just write the filename by itself:
% bali-phy ferns.fasta:{5-8,9-453,454-596,597-864,865-948,949-1328,1329-1393} --smodel gtr+Rates.gamma+inv --name=ferns-intron-exon --test
However, with partitioned analyses the command-line can become kind of long. Raw command lines are more useful with a simple unpartitioned analysis:
% bali-phy ferns.fasta --test
</para>
<para>You can see the options for bali-phy by running:
% bali-phy help
You can also get help on particular options or models with the help command:
% bali-phy help smodel // help on option
% bali-phy help gtr // help on model
</para>
</section>
<section><info><title>Running an analysis</title></info>
After you get the <userinput>--test</userinput> command to work, delete <userinput>--test</userinput> and start two copies of this analysis:
% bali-phy -c ferns-intron-exon.txt &
% bali-phy -c ferns-intron-exon.txt &
These should create directories called <filename>ferns-intron-exon-1</filename> and <filename>ferns-intron-exon-2</filename>.
Lets look inside one of these directories:
% ls ferns-intron-exon-1
% less ferns-intron-exon-1/C1.log // numerical parameter samples
% less ferns-intron-exon-1/C1.trees // tree samples
% less ferns-intron-exon-1/C1.P1.fastas // alignment samples for partition 1
% less ferns-intron-exon-1/C1.P2.fastas // alignment samples for partition 2
</section>
<section><info><title>Monitoring an analysis</title></info>
You can monitor an analysis in two ways. One way is to use the program <userinput>tracer</userinput>.
% tracer ferns-intron-exon-1/C1.log ferns/intron-exon-2/C1.log // you may need to double-click on Tracer and load log files via the menu.
We would like to see that the likelihood and posterior probability have stabilized. Another method of monitoring is just to
generate the summary report.
</section>
<section><info><title>Interpreting parameter names</title></info>
The parameter estimates have names that look like context1/value or context1/model:parameter. For example S1/gtr:pi[A] indicates
<variablelist>
<varlistentry><term>S1</term><listitem>The 1st substitution model</listitem></varlistentry>
<varlistentry><term>gtr</term><listitem>The GTR (general, time-reversible) nucleotide substitution model.</listitem></varlistentry>
<varlistentry><term>pi</term><listitem>The nucleotide frequencies</listitem></varlistentry>
<varlistentry><term>[A]</term><listitem>The frequency for A (adenine).</listitem></varlistentry>
</variablelist>
We can get information on what a parameter means by running <command>bali-phy help <replaceable>model</replaceable></command>:
% bali-phy help gtr
For more information on parameter names, see <link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="http://www.bali-phy.org/README.xhtml#output">the section on Output / Field names</link> in the Users Guide.
</section>
<section><info><title>Summarizing the results of a run</title></info>
% bp-analyze ferns-intron-exon-1 ferns-intron-exon-2
% firefox Results/index.html
Let's concatenate the summmary estimates for each partition to recover an alignment for the whole gene region.
% alignment-cat Results/P{1..7}-max.fasta > ferns-bali-phy.fasta
Since bali-phy is still running, we can get information on its progress by rerunning <userinput>bp-analyze</userinput>.
</section>
<section><info><title>Comparing to MAFFT, FSA, and PRANK</title></info>
We can see how the different aligners affect the total alignment length:
% alignment-info ferns-bali-phy.fasta | grep columns
% alignment-info ferns-prank.fasta | grep columns
% alignment-info ferns-fsa.fasta | grep columns
% alignment-info ferns-fsa0.fasta | grep columns
% alignment-info ferns-mafft.fasta | grep columns
Let's visually compare the alignment by loading all the alignments in aliview. Shrink each alignment so that it is very small, and look at the intron between exons 2 and 3.
% aliview ferns-bali-phy.fasta &
% aliview ferns-prank.fasta &
% aliview ferns-fsa.fasta &
% aliview ferns-fsa0.fasta &
% aliview ferns-mafft.fasta &
</section>
</section>
</article>
|
