nithins / proteinvis Goto Github PK

*******************************************************************************
Hardware and OS requirements

Discrete graphics card with support for Opengl 2.0 and GLSL 
GL_EXT_geometry_shader4 extensions.

Linux 2.6 with the drivers for the Gpu installed

*******************************************************************************

Dependant libraries

The following libraries and tools are needed to build the package
* qt 4.x -- qt gui lib (http://qt.nokia.com/)
* qt-dev-tools (qmake ,moc,uic)
* qt-dev-headers 
* qt-opengl-lib and dev headers
* opengl libraries and dev-headers (libGL,libGLU)
* libGLEW library and headers (http://glew.sourceforge.net/)
* boost regex library and headers (http://www.boost.org/)
* cmake 2.6 (http://www.cmake.org/)

Note:QT boost GLEW and cmake are usually available on standard linux distros.

*******************************************************************************
Building 

make sure that glapp_framework is available in the proteinvis source tree 
either via symlink or directly. (i.e there should be a directory called
proteinvis/glapp_framework containing the framework sources).

CMake reccomends that you build in a fresh clean directory and not in the
source tree itself

in the build directory type "cmake <path to proteinvis source tree>".

Note:: Sometimes you may need to run the above command again to update the 
cmake cache. 

type make .. if all goes well you should have protvis in the build dir.

*******************************************************************************
Running


The following is a list of the command line options.

* -pf [crd/pdb file]   -------- is always needed.without this app will not launch

* -sf [off file] -------------- surface mesh file in off format

* -acf [alp file] ------------- The output from the alpha complex code (Volume.tar)

* -tetf [tet] -pocf [tepoc]---- The output from the pocket calculation code. The
file is a dump of the delone tetrahedra .. 
First line specifies no. of tetrahedra 
Second comment on the order of info 
Third - End tets written as (i v1 v2 v3 v4).

Make sure you specify all data for a single protein. There isnt much code yet to 
check validity of the files and will read anything it is given.

For loading multiple proteins specify all data pertinent to one protein after the
-pf tag. The -pf switches act as delimiters for multiple protein file sets. 

eg: proteinvis -pf one.crd -acf one.acf -sf one.off -pf two.crd -pocf two.tepoc.all 
    -tetf tow.tet

*******************************************************************************

Render window commands.

Within the render window keystrokes and mouse behave in 3 broad modes

To switch b/w modes press ctrl-shift-'<'/ctrl-shift-'>' or alternatively click 
on the respective obj (for the framework click any where other than a viewed obj
the switch in mode will be communicated in the shell output.

---------Model viewing actions--------

Click and drag the MLB to move the model around. Click and drag the MRB to rotate 
the model. 

press ctrl and use the mouse wheel to resize the model in world space

---------Framework viewing actions-----

Press 'x' to provide you with a world space axes. (None for the model space yet).

press 'w' 'a' 's' 'd' to move front left back right respectively from the current 
viewing position. 

press 'r' to reset the view to its initial state. 

press 'b' to toggle back face culling

press 'B' to change the back face convention (clock wise or counter CW)

press 'L' to enable/disable lighting.

---------Light viewing actions---------

use the mouse to move/orient the light (same as controlling the model)

use the mouse wheel to make the graphic grow in size to allow easier control

use the mouse middle mouse button to turn off light.

---------Common actions-----------------

press 'ctrl+shift +<' to change the current reciever. 

press 'ctrl+shift +>' to change the current reciever. 

----------------------------------------

Note:Most changes Made to will be acknowledged in the status bar. 

Note: to make the pick operation, which is used to switch modes,more efficient, the 
model is rendered as a box (only for selection..:) ). Hence when some areas that are
visibly not part of the model are clicked will still register as a model click. 

Note: As part of QT's widget handling you will be required to click in the render 
window so that it may accept input. So if your buttons are not functioning this 
is probably whats wrong..:). 

*******************************************************************************
Windowing interface
===============================================================================
Model UI.
===============================================================================
the model ui is self-explanatory. Only those UI elements which have the necessary 
files specified in the command line while launching will be enabled. 

The color selection allows you to group the atoms by Atom(No grouping),Atom type,
Amino Acid,.... After you make your selection and set the colors hit the refresh
button to actually push the colors. For now the colors need to be specified in 
#RRGGBB format. Will implement multiple selection and colorpicker later. 

The buttons save transformation and load transformation save the current orientaion
positioning and sizing of the molecule to a .trans file. You could use one 
transformation to open another molecule though it makes most sense only when u 
want to save the state for the same molecule(s). When there are multiple molecueles 
there is an additional scaling to fit all the molecueles into the same frame. This 
transformation does not account for that. Also you'll have to save the state for
each molecuele separately for now.

===============================================================================
Framework UI
===============================================================================

Snapshot and record button. The image format is png. They will be stored in the 
directory from which the bin was launched. The record feature will create a dir 
and put the seq of images there the snapshot feature will just dump a .png file. 
The filename is random (it will not clash with any other file in the dir ..) 

===============================================================================
Light UI
===============================================================================
The framework allows for a maximum of 2 lights in the scence. ( More can be added
but really, not too many lights are needed for a good scene). 

For shiny surfaces raise the specular value near to white(both light and material). 
Similarly for diffuse surfaces (material diffuse and ambient is not enabled) raise
the diffuse value and reduce specularity. Raise the global specularity to allow 
seeing smaller shine area's (very coarsely speaking ..:) ). 

Use the disable light graphic to disable drawing the light graphic. 

*******************************************************************************