A set of tools for analyzing a collection of heteropolymer structures, and for measuring properties of or producing information about the ensemble as a whole. These were developed with the Rosetta software suite in mind; however, they may be more broadly useful, so I am first releasing this code under a permissive (MIT) licence. This allows it to be altered for incorporation into other software projects should it prove useful.
Vikram K. Mulligan ([email protected])
The patches/ directory contains the following patchfile(s), which can be used to add EnsembleMetrics to Rosetta. Note that the intention is for these to be incorporated into the public releases of Rosetta in the near future.
| Name | Description | Can be applied to |
|---|---|---|
| ensemble_metrics.patch | The EnsembleMetrics base class, plus the derived CentralTendencyEnsembleMetric (which measures mean, median, mode, etc. of an input value produced by a Rosetta SimpleMetric). |
Rosetta Git SHA 9907e74b22a2c9bb71d52b56a7110a107baf621d (master branch, 8 March 2022). |
| mpi_support.patch | Adds support for analysing a large ensemble of poses on a cluster using MPI (massive parallelism). | Rosetta Git SHA 9907e74b22a2c9bb71d52b56a7110a107baf621d (master branch + ensemble_metrics.patch, 11 March 2022). |
Patchfiles can be applied by navigating to your Rosetta/main directory (e.g. cd my_rosetta_installation/Rosetta/main), copying the patchfile to the current directory, and using the Linux patch command:
patch -p1 < ensemble_metrics.patchThe src/protocols/ensemble_metrics directory contains source code for the EnsembleMetrics framework. Although this is intended to be compiled against Rosetta headers and linked against Rosetta, one may easily replace Rosetta objects (such as Poses, SimpleMetrics, and ResidueSelectors) with their equivalents from other software packages. (The beauty of an MIT licence is that it allows full refactoring of the code to suit one's needs).
In src/protocols/ensemble_metrics, source code for derived classes (particular EnsembleMetrics) such as the CentralTendencyEnsembleMetric may be found. The src/protocols/init directory contains initialization functions for a factory system (which may or may not be useful in a new context). The src/protocols/parser directory contains code allowing the instantiation of EnsembleMetric subclasses when they are invoked in an XML script. (These functions are used to make EnsembleMetrics accessible to the RosettaScripts scripting language in Rosetta, but could be useful elsewhere.)
The test directory contains unit tests for the derived classes of the EnsembleMetric base class.
This work is currently unpublished. If you find it useful for your work, please consider including the author (Vikram K. Mulligan) in your publication. This page will be updated when this work is published.
This work is made available under an MIT licence (see LICENCE file). This gives anyone great leeway to modify, refactor, or redistribute this code, and to use it in their own personal, open-source, or commercial projects.