This file is indexed.

/usr/share/doc/coop-computing-tools/manual/sand.html is in coop-computing-tools-doc 4.0-2.

This file is owned by root:root, with mode 0o644.

The actual contents of the file can be viewed below.

  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
215
216
217
218
219
220
221
222
223
224
225
226
227
228
<html>

<head>
<title>SAND User's Manual</title>
</head>

<body>

<style type="text/css">
pre {
background: #ffffcc;
font-family: monospace;
font-size: 75%
font-align: left;
white-space: pre;
border: solid 1px black;
padding: 5px;
margin: 20px;
}
</style>
<h1>SAND User's Manual</h1>
<b>Last Updated January 2011</b>
<p>
SAND is Copyright (C) 2010- The University of Notre Dame.
This software is distributed under the GNU General Public License.
See the file COPYING for details.
<p>
<h2>Overview</h2>

<p>SAND is a set of modules for accelerating genome assembly and other
bioinformatics tasks.
Using the  <a href="http://www.cse.nd.edu/~ccl/software/workqueue/">Work Queue</a>
framework, SAND can distribute computational tasks to hundreds of nodes harnessed
from clusters, clouds, grids, or just the idle machines you have in your office.
SAND can be used as a drop-in replacement for the conventional overlapper in
the Celera Assembler, or can be used as a standalone tool by advanced users.

<p>SAND is part of the <a
href=http://www.cse.nd.edu/~ccl/software>Cooperating Computing
Tools</a>.  You can download the CCTools from <a
href=http://www.cse.nd.edu/~ccl/software/download>this web page</a>,
follow the <a href=install.html>installation instructions</a>, and you
are ready to go.</p>

<h2>Using SAND with the Celera Assembler</h2>

If you are already using the Celera Assembler version 5.4, you can easily switch to using SAND
to accelerate assemblies by using our modified <tt>sand_runCA_5.4</tt> script. If you are using the newer 
Celera Assembler version 6.1, you can switch to SAND by using the <tt> sand_runCA_6.1 </tt> script. If you are using 
Celera Assembler version 7.0, you can switch to SAND by using the <tt> sand_runCA_7.0 </tt> script. We assume that you have
installed SAND according to the instructions above and addeed it to your <tt>PATH</tt>.

<ol>
<li> Copy the script <tt>sand_runCA_X.X</tt> into the same directory where you have the normal <tt>runCA</tt> installed.
<li> Set <tt>ovlOverlapper=sand</tt> to your <tt>spec</tt> file.
<li> For Celera 7.0, add <tt>sandAlignFlags=-e "-o ovl_new"</tt> to your <tt>spec</tt> file. This tells the alignment program
to output new Celera 7.0 style overlaps.
<li> Run <tt>sand_runCA_X.X</tt> just like you normally use <tt>runCA</tt>
</ol>

You will see the assembly start as normal.  When the overlapping stage begins, you will see output like this:

<pre>
 Total | Workers   | Tasks                      Avg | Candidates
  Time | Idle Busy | Submit Idle  Run   Done   Time | Found
     0 |    0    0 |      0    0    0      0    nan | 0
     5 |    0    0 |      0    0    0      0    nan | 0
    10 |    0    0 |      0    0    0      0    nan | 0
    15 |    0    0 |      0    0    0      0    nan | 0
...
</pre>

SAND is now waiting for you to start <tt>work_queue_worker</tt> processes.
Each worker that you start will connect back to the master process
nd perform small pieces of the work at a time.
In general, the more machines that you can harness, the faster the work will go.
<p>
For testing SAND, just open a new console window and start a single worker,
specifying the hostname where the master runs and the port number it is listening on.
(This can be changed with the sandPort option in the CA spec file.)
For example:

<pre>
% work_queue_worker master.somewhere.edu 9123
</pre>

With one worker, you will see some change in the output, like this:
<pre>
 Total | Workers   | Tasks                      Avg | Candidates
  Time | Idle Busy | Submit Idle  Run   Done   Time | Found
     0 |    0    0 |      0    0    0      0    nan | 0
     5 |    0    1 |    100   83    1     16   0.32 | 1858
    10 |    0    1 |    100   69    1     30   0.33 | 3649
    15 |    0    1 |    100   55    1     44   0.34 | 5464
</pre>

For a very small assembly, a single worker might be sufficient.
However, to run a really large assembly at scale, you will need to
run as many workers as possible.  A simple (but tiresome) way of doing
so is to <tt>ssh</tt> into lots of machines and manually run <tt>work_queue_worker</tt>
as above.  But, if you have access to a batch system like
<a href=http://www.cs.wisc.edu/condor>Condor</a> or
<a href=http://www.sun.com/software/sge>SGE</a>, you can use them
to start many workers with a single submit command.
<p>
We have provided some simple scripts to make this easy.
For example, to submit 10 workers to your local Condor pool:
<pre>
% condor_submit_workers master.somewhere.edu 9123 10
Submitting job(s)..........
Logging submit event(s)..........
10 job(s) submitted to cluster 298.
</pre>

Or, to submit 10 worker processes to your SGE cluster:
<pre>
% sge_submit_workers master.somewhere.edu 9123 10
Your job 1054781 ("worker.sh") has been submitted
Your job 1054782 ("worker.sh") has been submitted
Your job 1054783 ("worker.sh") has been submitted
...
</pre>

Note that <tt>condor_submit_workers</tt> and <tt>sge_submit_workers</tt>
are simple shell scripts, so you can edit them directly if you would
like to change batch options or other details.
<p>
Once the workers begin running, the SAND modules can dispatch tasks to
each one very quickly.  It's ok if a machine running a worker crashes
or is turned off; the work will be silently sent elsewhere to run.
<p>
When the SAND module's master process completes, your workers will
still be available, so you can either run another master with the
same workers, remove them from the batch system, or wait for them to
expire.  If you do nothing for 15 minutes, they will automatically exit.

<h2>SAND in More Detail</h2>

This section explains the two SAND modules in more detail,
if you would like to tune the performance or use them independently
of Celera.  We assume that you begin with a file of reads in FASTA format.
To use the SAND modules, you must first generate repeats and compress the data.  (If you don't have data handy, download <tt>small.cfa</tt> and  <tt>small.repeats</tt> data from the <a href=http://www.nd.edu/~ccl/software/sand>SAND Webpage</a>.)

To generate repeats from a FASTA file, use the <tt>meryl</tt> tool from the Celera Assembler:
<pre>
% meryl -B -m 24 -C -L 100 -v -o small.meryl -s small.fa
% meryl -Dt -s small.meryl -n 100 > small.repeats
</pre>

Then use <tt>sand_compress_reads</tt> to compress the sequence data into a compressed FASTA (.cfa) file:
<pre>
% sand_compress_reads small.fa small.cfa
</pre>

The filtering step will read in the compressed sequence data (<tt>small.cfa</tt>) and quickly produce a list of candidate sequences (<tt>small.cand</tt>) for the following step to consider in detail.  Start the filtering step as follows:
<pre>
% sand_filter_master -r small.repeats small.cfa small.cand
</pre>

While the filtering step runs, it will print some statistics to the
console, showing the number of workers available, tasks running, and so forth:

<pre>
 Total | Workers   | Tasks                      Avg | Candidates
  Time | Idle Busy | Submit Idle  Run   Done   Time | Found
     0 |    0    0 |      0    0    0      0    nan | 0
     5 |    0   14 |    158   14   14    130   0.39 | 16452
    10 |    0   15 |    356   15   15    326   0.38 | 42382
    15 |    0   15 |    549   15   15    519   0.38 | 69055
    20 |    0   15 |    744   15   15    714   0.38 | 96298
    25 |    0   15 |    942   15   15    912   0.38 | 124284
</pre>

The alignment step will take the list of candidates generated in the previous step (<tt>small.cand</tt>),
the compressed sequences (<tt>small.cfa</tt>) and produce a listing of how and where the sequences
overlap (<tt>small.ovl</tt>).  For example:

<pre>
% sand_align_master sand_align_kernel -e "-q 0.04 -m 40" small.cand small.cfa small.ovl
</pre>

The options <tt>-q 0.04 -m 40</tt> passed to <tt>sand_align_kernel</tt> indicate a minimum alignment quality of 0.04 and a minimum alignment length of 40 bases.
Again, a progress table will be printed to standard out:
<pre>
 Total | Workers   | Tasks                      Avg | K-Cand K-Seqs | Total
  Time | Idle Busy | Submit Idle  Run   Done   Time | Loaded Loaded | Speedup
     0 |    0    0 |      0    0    0      0   0.00 |      0      0 |  0.00
     8 |    0   48 |    100   52   48      0   0.00 |   1000    284 |  0.00
    10 |    0   86 |    100   13   86      1   7.07 |   1000    284 |  0.71
    36 |    1   83 |    181   14   83      2  19.47 |   1810    413 |  1.08
   179 |    1   83 |    259   92   83      3  22.51 |   2590   1499 |  0.38
   186 |    2   80 |    259   15   80     85  28.54 |   2590   1499 | 13.04
   199 |    2   80 |    334   90   80     86  29.96 |   3340   1499 | 12.95
   200 |    2   80 |    334   90   80    114  59.43 |   3340   1499 | 33.88
   202 |    2   81 |    334    9   81    165  86.08 |   3340   1499 | 70.32
</pre>

After the sequence alignment step completes, you will have an overlap (<tt>.ovl</tt>)
file that can be fed into the final stages of your assembler to complete the consensus step.

<h2>Tuning Suggestions</h2>

<dir>
<li> As a rule of thumb, a single task should take a minute or two.  If tasks are much longer than that, it becomes more difficult to measure progress and recover from failures.  If tasks are much shorter than that, the overhead of managing the tasks becomes excessive.  Use the <tt>-n</tt> parameter to increase or decrease the size of tasks.
<li> When using banded alignment (the default), the <tt>-q</tt> match quality parameter has a significant effect on speed.  A higher quality threshhold will consider more alignments, but take longer and produce more output.
<li>
The columns of the output are as follows:
<dir>
<li> Total Time is the elapsed time the master has been running.
<li> Workers Idle is the number of workers that are connected, but do not have a task to run.
<li> Workers Busy is the number of workers that are currently running a task.
<li> Tasks Submitted is the cumulative number of tasks created by the master.
<li> Tasks Idle is the number of tasks waiting for a worker.
<li> Tasks Running is the number of tasks currently running on a worker.
<li> Tasks Done is the cumulative number of tasks completed.
<li> Avg Time is the average time a task takes to run.  An average time of 60 seconds is a good goal.
<li> K-Cand Loaded indicates the number of candidates loaded into memory (in thousands).
<li> K-Seqs Loaded indicates the number of sequences loaded into memory (in thousands).
<li> Speedup is the approximate speed of the distributed framework, relative to one processor.
</dir>

</dir>

<h2>For More Information</h2>

For the latest information about SAND, please visit our <a href=http://www.cse.nd.edu/~ccl/software/sand>web site</a> and subscribe to our <a href=http://www.cse.nd.edu/~ccl/software>mailing list</a>.
</body>
</html>