JalView 1.3 beta - Michele Clamp 1998
-------------------------------------

Any problems mail me:  michele@ebi.ac.uk

--------------------------------------------------
Note: 18th September 1998:  New in 1.3b version

  Grouping sequences by selecting on the average PID tree
  Conservation analysis and colouring by group
  Sequences can be sorted by tree and group

  New colour schemes - K&D hydrophobicity, 
                       Helix,sheet,turn,buried propensities

  Can fetch sequence features and display on the alignment

  Can view database entry through the application

  Can save to a file for the text alignment and postscript.

  New command line options: -srsServer for the location of wgetz 
                            -database for srs database

  New applet parameters:  srsServer -  for wgetz
                          database  -  for srs database name

   (n.b. srsServer should be of the form
             srs.ebi.ac.uk/srs5bin/cgi-bin/
    i.e. no http:// and must have a following slash
   )
                           
--------------------------------------------------
Note: 17th August 1998:  New in 1.2b version

 In summary:
  User defined colour schemes
  Tree calculation and drawing.
  Principal component analysis to cluster 
         sequences.
  Extra fonts and 2 drawing modes.

 In more detail:

  New file formats accepted : BLC and MSP
  (output of MSPcrunch for blast output viewing).

  Selecting columns and trimming the alignment
  left and right of selected columns.  This 
  is very useful for evening up ragged edges
  of an alignment prior to calculating a tree
  or PCA.

  Gap insertion character can be changed from
  - to .

  Extra fonts - Helvetica and Times-Roman on the
  screen display and the choice of plain or bold
  fonts.

  Two modes for font drawing.  Fastdraw 
  only allows Courier but is the fastest in
  updating the screen and has the residues
  closer together.  Non-fastdraw allows Courier,
  Helvetica and Times and spaces the residues
  further apart from each other.

  Selecting a colour scheme or a PID threshold 
  from the Colour menu only changes the colours
  of the selected sequences.  If none 
  are selected then the changes affect all
  sequences.

  Extra colour schemes.  I have included Willie
  Taylor's colour scheme (the subject of a very
  entertaining Protein Engineering paper vol 10, 
  743-746 1997) as an alternative to zappo.  

  User definable colour schemes. Once 
  a colour scheme is set you can change it by 
  selecting the 'User defined colours...' option
  in the Colour menu.  You can change the colours
  of the different residue groupings by using a
  rgb colour selector.  You can also change which
  residues are grouped together. (Note: there
  are problems here if two colours are set equal).

  The PID colours have changed from blues to 
  red pink and yellow for no particular reason.

  You can sort sequences in the alignment by their
  percent identity to the consensus sequence.

  You can remove redundancy in the sequences by 
  percent identity to the consensus sequence.

  You can sort the sequences by group for tidiness.

  Using the whole sequences as the starting data 
  (as opposed to the individual residues) you can
  perform a principal component analysis of your
  alignment.  This is a method of clustering 
  aligned protein sequences.  For more information
  have a look at http://industry.ebi.ac.uk/SeqSpace
  or Georg Casari, Chris Sander and Alfonso Valencia. 
  Structural Biology volume 2, no. 2, February 1995.
  The sequences are converted into vectors (n dimensions
  for n sequences in your alignment).  These
  vectors can then be viewed in 3d by projecting
  them into a lower dimensional space. 

  Tree building using percentage pairwise
  identity of aligned sequences as the distance matrix.
  The trees are then built using either UPGMA
  or neighbour joining.  Postscript output can
  be obtained by email.

  Clustalw alignment locally or at the EBI. Choosing one of
  these options with either run clustalw locally or
  send your sequences to the EBI for alignment.  The
  resulting aligned sequences are displayed in a new window.
  The local option only works unded unix at the moment.
 
  Selecting sequences in an alignment will also select
  the same sequences in a tree or PCA window.

  Local alignment now works with windows 95 IF
  clustalw is in the path and c:\windows\temp is
  writable.

  added -mail mailserver option to command line

--------------------------------------------------
Note: 5th June 1998:  This is unfinished software
that I am making available to anybody who might
like it with the usual disclaimer that I take
no responsibility for anything at all that may
go wrong with it.  I will be extremely grateful
for any suggestions/gripes anybody may have.
If the load becomes too much for my pine INBOX
I may set up a mailing list for software updates,
bug fixes etc.
--------------------------------------------------

This is a pre-beta release  of the java multiple alignment
editor Jalview . 

To install Jalview look in the INSTALL file.

(Quick start if you have java already set up and 
you know what you're doing:

Unix:
    setenv CLASSPATH ${CLASSPATH}:`pwd`/jalview.jar
    java jalview.AlignFrame examples/lipase.msf File MSF

PC:
    set CLASSPATH=%CLASSPATH%;jalview.jar
    java jalview.AlignFrame examples\lipase.msf File MSF

}


##########################################################

PLEASE PLEASE PLEASE if you have problems don't just say
'Oh bugger it' and stroll off for a cup of coffee.  
Please take some time to email me so I can put things 
right.  Any feedback is welcome but  don't make it
too negative. We're delicate little flowers here at
the EBI.

##########################################################


For help on writing html pages to run Jalview look in
the doc/applet.html file.

For general information about Jalview look in the
doc/contents.html file.

For details of the functions in Jalview look in the
doc/help.html file.

You can download this distribution (and later ones)
from http://www.ebi.ac.uk/~michele/jalview/dist/

Jalview's home page is :
  http://www.ebi.ac.uk/~michele/jalview/

Latest releases will appear on the home page.


New in 1.1b July 1998
--------------------

Can run clustalw locally.  Clustalw needs to be in your path and named
'clustalw' otherwise things go pearshaped.  You can also run clustalw
remotely at the EBI.  This is new stuff and probably has hideous bugs.

Principal component analysis of alignments.  Reduces sequences in your
alignment to points in a 3d space so you can see how the sequences
cluster.  This is meant to be complementary to constructing a tree from
your alignment.

Colour your sequences only if the residues agree with the consensus above
a certain percentage.  This feature is available with zappo and
hydrophobicity colourings.

Trimming alignments on the left and right hand side (again stolen from
bevu). Selecting a column on the scale panel and choosing 'remove columns
left of selected column' from the edit menu deletes all columns left of
the selected one.  I find this useful for evening up the edges of
alignments prior to PCA or making a tree as uneven sequence lengths bias
the output.

umm...changing gap insert characters from - to . and vice versa.

Extra editing functions - select all, invert selected sequences..

If no sequences selected and the consensus option is picked a consensus is
calculated over all sequences.  Previously an error message came up in the
status bar saying no sequences selected.

