Information Technology Reference
In-Depth Information
3.
User interfaces for EMBOSS
Most EMBOSS programs will run in a text only environment, such as can be achieved
using telnet or ssh to connect to an EMBOSS UNIX server such as provided by the
(RFCGR). However, some programs generate graphical output which requires a terminal
with graphical capabilities. This can be achieved by running an X server (such as
Hummingbird eXceed for Windows) on the terminal, or by using a suitable EMBOSS GUI
(such as the web GUI
W2H
or the Java GUI
Jemboss
).
3.1
EMBOSS from the UNIX command line
The most common method of organising files and folders on a UNIX machine is to type
textual commands in response to a UNIX prompt (
unix%
, say) forming a UNIX command
line. It is possible also to run programs (including those of the EMBOSS package) by
forming command lines. A command line to invoke an EMBOSS program specifies the
program to run, together with all parameters for the specific analysis. The command line
starts with the name of the application, followed generally by the input files.
Simply pressing <return> will set the program running. The researcher will be
prompted for any other information necessary for the program to function, including the
name of a file in which textual output should be saved. The application will then run, and a
UNIX prompt returned to the screen. The results are now contained in the output file, and
must be investigated separately. If further options on the program need to be accessed, the
-
opt
qualifier
3
may be added onto the command line. This indicates to the program to
prompt for further options.
There are some options that a program will never prompt for, and need always to be
written on the command line. For each program, these are found in the EMBOSS help
section under the heading “Advanced Qualifiers”.
One of the major computational features of EMBOSS, is how it finds sequence files
to read in. The Uniform Sequence Address (USA) is the form that the file address must
adhere to in order for the program to function. This is generally of the type
format::database:file entry
. A similar address may be used for input or output files.
Generally the format section can be omitted for input sequences, as EMBOSS should
automatically recognise this. This is not the case if the file is in raw format, or
Intelligenetics format.
If a multiple sequence alignment is required, input on the command line must be a
single file, or a set of wild-carded file names. It is not always the case that this file contains
multiple sequences, it may also contain database and accession numbers, or files names
within a folder. Such a file is known as a “list file” or “file of filenames”, and must be pre-
pended with an @ sign (or list::) in order to be recognised by the EMBOSS application.
Documentation on each of the programs in EMBOSS may be invoked by running
the program
tfm
. The program got its name from rather exasperated developers bombarded
with questions from users who had not read The Fine4 Manual. It has been written for both
users and developers and so may be biased towards computer jargon.
To access the information, users must type
tfm
on the command line, plus the name
of the application they need help on. Alternatively, all this documentation may be found on-
line at: http://emboss.sourceforge.net/apps/
3
A qualifier alters the behaviour of the original program. EMBOSS programs have many of these, to cater for
the vast number of options needed to perform specific analyses.
