BOXSHADE - List of frequently asked questions
Q: What is BOXSHADE ?
BOXSHADE is a program for pretty-printing multiple alignment output. The program itself doesn't do any alignment, you have to use a multiple alignment program like ClustalW or Pileup and use the output of these programs as input for BOXSHADE. Of course, you can also use manually aligned sequences (in a proper format, see below).
Q: What is the most recent version ?
The latest release version is 3.2, a new version (3.21) with slight improvements in RTF output will be made available soon.
Q: What input formats can be read by BOXSHADE ?
MSF format as written by PILEUP, READSEQ, or SEQIO (fmtseq).
ALN format as written by CLUSTALW.
Other common formats have to be converted to MSF first (by READSEQ or a similar program)
Q: What output formats can be written by BOXSHADE ?
ASCII plain text files (showing either identical or different residues)
PS and EPS format (using background shading and/or colors)
HPGL format (using colors)
RTF format (using background shading and/or colors)
XFIG format (using background shading and/or colors)
PICT format (using background shading and/or colors).
Q: How does the algorithm work?
For some more details see the documentation file that comes with the BOXSHADE program or see BOX_doc.html
A brief description: In the first step, the program checks for every column in the alignment, if there are enough identical/similar residues to form a consensus (threshold can be defined by user).
In the second step, the program checks for every residue if it is identical or similar to the consensus of the corresponding alignment column. If there is no such consensus or the residue is neither identical nor similar to the consensus, this residue will be printed in normal rendition. If the residue is identical to the column-consensus, it will be printed in a special rendition (user-defined, usually inverse). If the residue is not identical but at least similar to the column-consensus, it will be printed in another user-defined rendition (usually on gray background). Positions in the alignment which are totally invariant can be treated specially.
Residue similarity is defined by a (editable) parameter file, not by any substitution matrix.
Q: Does it also work with DNA sequences?
Yes, there are also parameter files designed for DNA comparisons.
Q: On what platforms does BOXSHADE run ?
Currently, it runs on VMS systems (both VAX and AXP), on OSF1/Digital unix systems, and on MSDOS-based PCs (Win95 in DOS-boxes).
The source code is freely available, but it is written in PASCAL and therefore not easily portable. Since BOXSHADE was designed to be compiled under DEC-Pascal and Borland/Turbo-PASCAL, two very different dialects with no consistent method of conditional compilation, it frequently uses some pseudo-PASCAL statements with are resolved by machine-specific include files.
I have heard rumours that somebody compiled a SunOS/Solaris version but I don't know any details.
Q: Are there servers for BOXSHADE ?
There is a WWW-server running here at the ISREC in Lausanne/Switzerland using the URL: BOX_form.html
I don't know of any mirrors in the US. There is no mail server.
Q: Which versions of BOXSHADE are available, where do I get them from ?
It is available in the following versions:
Source code distribution
Precompiled executables for OSF1/Dec-Unix and VMS(AXP)
Currently, there are no precompiled versions of release 3.2 for PCs since we don't have access to a PC equipped with a PASCAL compiler. If anybody would volunteer to compile the source code, we would be very grateful!
Q: Who has written BOXSHADE / where can I report bugs ?
Boxshade has been written by Kay Hofmann and Michael D. Baron. The addresses are given below
Kay Hofmann, E-mail: email@example.com
Michael D. Baron Institute for Animal Health Ash Road Pirbright, Surrey GU24 0NF U.K.
Please report bugs only when you are sure that you really did everything by the book, after reading the documentation and checking the FAQ list for known problems. I currently get much more BOXSHADE-related mail than I can handle, that is why I compiled this FAQ list. Michael Baron should be contacted only for problems which are specific for PICT output. Kay Hofmann should be contacted for all other problems, including problems with the WWW-server.
Q: How can I cite BOXSHADE / Is there a publication on BOXSHADE ?
There is no publication on BOXSHADE and none is planned. Most people just use it for figures in publications and don't mention anything, this is ok for us. If you really feel like mentioning BOXSHADE, you could either acknowledge it in the figure legend or in the Mat&Meth part on sequence analysis.
Problems with specific input formats
Q: Why does my MSF file not work with BOXSHADE?
BOXSHADE does not work with MSF files written by ClustalW. When working with ClustalW, use the .ALN files (Clustal's native format) as output.
Q: My input format is not recognized, what can I do?
Most common formats for multiple sequences can be converted to MSF format by READSEQ or the SEQIO package.
Q: Is there a difference between dots or dashes as gap-symbols?
You may use either. Both are regarded as gaps. They will remain dots or dashes in the output files.
Problems with specific output formats
Q: What can I do with my postscript (PS/EPS) output ?
POSTSCRIPT format is usually meant for printing it directly on a laser printer with PS-capabilities.
On UNIX computers, just send the output (.ps) file to the printer using the lp or lpr command.
On PCs under DOS, send the files to the printer using a command like
copy output.ps lpt1:
but make sure that your printer really understands POSTSCRIPT. (See also next question.)
On Macs, there are freeware utilities for sending POSTSCRIPT files to the printer (all Apple Laserwriters understand POSTSCRIPT)
POSTSCRIPT files can be previewed on screen by freely available PS-previewers like ghostscript/ghostview and psview. Ghostscript is also very useful for printing POSTSCRIPT output on non-POSTSCRIPT printers (dot-matrix, inkjets, cheap laser printers).
EPS files are meant for inclusion in other documents. The can be imported by many word processors and graphics programs but can not be edited.
Q: When printing POSTSCRIPT output on my printer,
it produces tons of garbage!
Two problems could be causing that:
Check if your printer really understands POSTSCRIPT.
PS files created by older BOXSHADE versions sometimes have problems with the automatic format recognition of multi-mode printers like HP-laserjets of the 'M-series'. You should either get a more recent version of BOXSHADE or you have to edit the first line of the output file manually. For proper recognition, it has to look like the following line:
(other version numbers are also ok). Old BOXSHADE output files just have
which is sufficient for some printers but not for all of them.
Q: Is there any output format that can be edited
by a word processor or graphics program?
For Unix-workstations, XFIG can be used (see below).
For import to graphics programs running on Macs and PCs, the PICT format can be used. It can be read by almost every program. Big alignments cause problems, see questions below.
RTF files can be read and edited by most word processors. However, almost none of these is able do display individual character background. To my knowledge, the only word processor capable of doing that is MS-Word 7.0 . Since MS-Word 7.0 does this in a non-standard way, we have changed the RTF output routine in BOXSHADE V3.21 to support MS-WORD 7.0.
Q: Why does my computer choke on these PICT files?
BOXSHADE treats every single character and every background box as a drawing object. A typical alignment as a PICT file uses a huge amount of memory and cpu cycles. Use the PICT option only for relative short alignments if you need to insert labels or other features.
Q: Why does my RTF output look garbled?
The RTF files written by BOXSHADE contain only minimal formatting information. They usually require some manual post-processing before printing them. Typically, the page orientation and margins have to be set properly, it may also be necessary to chose a different (smaller) font for the alignment to fit on the page. Only use fixed-width fonts for the alignment !
Q: My word processor does not display background shading in RTF output!
As mentioned above, the only word processor I know to support individual character background is MS-Word 7.0. RTF has a specification for character background but this field is usually ignored by most word processors. Note that even MS-Word 7.0 does character background in a non-standard way (using the 'highlight' feature). Only versions of BOXSHADE starting from 3.21 can generate 'highlighted' RTF-files.
Q: What is XFIG output good for ?
xfig is a freeware drawing program for Unix/X-windows computers. It uses a relatively simple file format (in pure ASCII text) that can be also edited manually. BOXSHADE's XFIG output can be used by xfig versions above 2.1
Specific Problems with the stand-alone program
Q: Why does BOXSHADE give error messages with long input files ?
BOXSHADE has some built-in limitations for the maximum number of sequences and the maximum length of the alignment. These limitations are critical only with the DOS-version. The default limits for the DOS version are 20 sequences with 1300 residues each. By recompiling the source code it is possible to increase one of these parameters while decreasing the other. However, since we currently don't have access to PASCAL-equipped PCs, we are not able to provide those versions.
Specific problems with the WWW-server
Q: What does "404 Not Found" mean?
This cryptic error message means that no output file has been found since it has not been generated. The usual reason is that the input file was not recognized by the BOXSHADE server. Read the document box_error.html for a description and examples for valid formats.
Q: Why is my output truncated ?
PICT output is generally restricted to one page. I received a few complaints from people getting truncated output with different formats. Up to now I was never able to reproduce this problem when trying it myself, so I assume that these users probably did something wrong in the cut & paste process
(remember: Always use complete MSF or ALN files, never try to work with snippets from those files ! ).
Q: When I download the output, my web browser displays stupid things on the screen, what can I do?
The WWW-browser (like Mosaic or Netscape or Internet Explorer) should not try do display the BOXSHADE output on screen (except if a appropriate helper application is installed). Instead, the files should be downloaded to a local file on your computer for further processing. Normally the server should do this automatically. In some cases, it might be necessary to tell your browser explicitly to download the file instead of displaying it.
Q: The Server does not recognize my input format, what can I do?
If you have a non-standard MSF files (maybe written by Clustal) or you are not sure if you have the correct number of blank lines between the alignment blocks, use the option 'other format' instead of 'MSF'. When doing this, the server applies the READSEQ reformatting program first. This trick should also work for PHYLIP or PEARSON-formatted files.