March 2009 - IMP-dev - salilab.org

exposing MolecularHierarchyDecorators to python
by Keren Lasker 18 Mar '09

18 Mar '09

hi, I am trying to use write_pdb, which gets as input a set of MolecularHierarchyDecorator (MolecularHierarchyDecorators) This type (MolecularHierarchyDecorators) is not exposed in python. I tried to add typedef in the atom.i file, but then the compiler complained that this has already been declared in MolecularHierarchyDecorator.h ( which is true). but still the python interface does not include MolecularHierarchyDecorators. any ideas how to solve it ? thank you, Keren.

2 1

ParticleRefiner->Refiner
by Daniel Russel 18 Mar '09

18 Mar '09

To shorten the names (and since it was trivial) ParticleRefiner has become Refiner. If you use them explicitly anywhere, run this to update your code: sed -i.old "s#ParticleRefiner#Refiner#g" `find * -name \*.py -or -name \*.cpp -or -name \*.h` A better name would be appreciated and any suggestions are welcome.

1 0

Proposal: rename PrincipleComponents to PrincipalComponents
by Ben Webb 17 Mar '09

17 Mar '09

Unless the code is referring to some kind of PCA with which I'm not familiar, this should be Principal not Principle, surely? If no objections, I'll make the change. But obviously the interface will change when I do - just letting you know... Ben -- ben(a)salilab.org http://salilab.org/~ben/ "It is a capital mistake to theorize before one has data." - Sir Arthur Conan Doyle

2 1

Last week in IMP
by Daniel Russel 16 Mar '09

16 Mar '09

It was a quite week. - The new web site has gone live. Please check it out and pass on any comments/problems. - Several new geometric shapes have been added to the algebra module. - There is now a wiki page <https://salilab.org/imp/wiki/changes> which lists significant changes that have been made to IMP. Please update the page when you make major changes (the standard of what counts as major will emerge as we go on :-)

1 0

Interacting particles and score states
by Daniel Russel 12 Mar '09

12 Mar '09

The graph of interacting particles returned by calling get_interacting_particles() on all the restraints doesn't reflect the true interactions as soon as you involve things like centroids, rigid bodies or covers. One solution would be to allow ScoreStates to produce lists of interacting particles as restraints currently do (and have Domino add these lists to its graph). Other thoughts? Keren?

2 2

is there an easy way to copy a hierarchy ?
by Keren Lasker 12 Mar '09

12 Mar '09

I would like to implement a function copy_molecule ( which requires creating new particles with the appropriate attributes). More generally, I would like to have a functionality of taking a root of some hierarchy and copying all attributes values into a new root. did someone do something similar already ? thanks, Keren.

2 1

math functions in c++
by Daniel Russel 12 Mar '09

12 Mar '09

You need to use the std:: versions of the math functions when writing C ++ code. That is use "std::abs", "std::sin" etc. By using the C++ versions, rather than the C version, you make sure that the correctly overloaded version is used. If you, for example, use just "abs", then the argument is implicitly converted to an int, which is probably not what you want. Likewise, using "sinf" in algebra where we are using doubles for computation truncates the argument, loosing accuracy for no reason.

1 0

set_print_exceptions
by Keren Lasker 11 Mar '09

11 Mar '09

IMP.set_print_excpetions() does not exist anymore, at least in the swig interface - was the function renamed? thanks, keren.

2 3

xml support in imp
by Keren Lasker 06 Mar '09

06 Mar '09

hi all, I would like to add XML support in IMP. It seems that Arabica could be a good alternative - thoughts ? Keren.

3 2

Simplifying IMP usage
by Daniel Russel 06 Mar '09

06 Mar '09

So far development in IMP proper has mostly focused on expanding the capabilities and flexibility of IMP. The result is a set of functionalities which can be (and has to be) combined to get what you want done. At this point we should be starting to have a good idea what common usage patterns are and think about adding higher level functions which make implementing the usage patterns easier. These higher level functions should help narrow the gap between experimental data and IMP. Currently implemented examples of such things include: - copying a molecular hierarchy - setting up a rigid body defined by a set of particles - setting up a particle as a centroid or cover of a set of particles Future examples might include: - creating a molecular hierarchy for a representation of a protein with a given number of particles based on the number of amino acids in the protein - creating a rigid representation of a protein with a given number of spheres from a molecular hierarchy loaded from the pdb - building a hierarchical representation of a protein from a molecular hierarchy loaded from the pdb - setting up efficient volume exclusion restraints for an arbitrary collection of particles, rigid bodies and hierarchies. Anyway, I set up a wiki page for people and refine ideas and requests: <https://salilab.org/imp/wiki/high_level_operations>

1 0