September 2009 - IMP-dev

Search does not work for IMP API help
by Javier Ángel Velázquez Muriel 14 Sep '09

14 Sep '09

Hi, This useful feature is not working these days. At least Keren has the same problem. Anybody else? I've tried Firefox 3.5 and Safari, both Mac and Linux. Any help? Thanks! Javi

2 1

imppy.sh
by Javier Ángel Velázquez Muriel 14 Sep '09

14 Sep '09

Hi The line that sets PYTHONPATH only sets one directory right now. At least for me goes: PYTHONPATH='/chime1/home/javi/imp/trunk/build/lib' Probably is more useful with PYTHONPATH='/chime1/home/javi/imp/trunk/build/lib':$PYTHONPATH Javi

2 1

imp-deve meeting tomorrow
by Daniel Russel 07 Sep '09

07 Sep '09

Lets discuss release-related issues to get things back on track. 1:30.

1 0

get_interacting_particles
by Keren Lasker 07 Sep '09

07 Sep '09

hi, quick question: Given two proteins, represented as rigid bodies. Constructing a PairsRestraint on the two rigid bodies, what should be the result on the get_interacting_particles() ? for some reason I am getting 0, also after evaluating the restraint. I would have expected to get one list with two entries ( the two rigid bodies). what am i doing wrong ? thank you, Keren.

2 5

naming geometries
by Keren Lasker 05 Sep '09

05 Sep '09

hi, What is current status with naming of particles? Lets say I would like to name geometries in gs = IMP::display::XYZRsGeometry , which are lets say a few spheres. there is a set_name function to a specific geometry, but not sure there is a way to iterate geometries in gs. Seems it would have been easier if Geometry::set_name would use the name of a particle, but I could not find a say to do so anymore .... thank you, Keren.

2 1

This week in IMP
by Daniel Russel 04 Sep '09

04 Sep '09

- A new score IMP::core::BoundingBox3DSingletonScore has been added to keep particles in a box. - str should now work better in python when applied to IMP::Object objects and a few others. Objects should use the new IMP_SHOWABLE and IMP_SHOWABLE_INLINE macros instead of declaring show functions directory to make it so that they work in python. It is now a bug for print or str to not return something useful in python for any IMP object.

1 0

Getting the release back on track
by Daniel Russel 04 Sep '09

04 Sep '09

The release has slipped down in everyone's priorities, so we need to work on getting things back on track. There are a lot of entries in the schedule which say things will be done in may which, to my knowledge, are not done. Could people please check and update things. In addition, people need to make a big sweep through their code to make sure all functionality is either documented clearly or removed from the API. Some new things added to the list are - the IMP.search module does not appear to be used and is not well documented. I propose removing it from the release/build (make it an external module like GSL) - the IMP.statistics module, likewise has no non-trivial documented functionality. I propose likewise removing it from the release/build (by making it an external module like IMP.gsl). Thanks very much.

2 1

Proposed changes
by Daniel Russel 02 Sep '09

02 Sep '09

1) Change the interpretation of a RigidMember to simply be any particle which is moved rigidly with the rigid body, rather than a particle which is necessarily part of the representation used for collision detection and things. For example, you could then create a rigid body from a protein. This rigid body would have RigidMembers which are pieces of the representation of the protein at a variety of resolutions, as well as, possibly, markers for key parts of the protein. The RigidClosePairs finder would then take a Refiner which could be used to tell it that, for example, you want to perform collision detection between the leaves of the atom.Hierarchy. This change would require small changes to code which uses the RigidClosePairsFinder. 2) An a Object attributes to Particles which stores a pointer to an IMP::Object. This can be used for associating arbitrary data with the Particle. For example, Javi has images that he needs to associate with particles and I have collision detection hierarchies which can be used to accelerate finding close pairs or the closest pair between two rigid bodies. I'm tempted to merge these attributes with the Particle attributes, but am not sure of the performance implications. Comments?

1 0