![](https://secure.gravatar.com/avatar/cfe8857a24d561e8e700ab18e3ba7ec8.jpg?s=120&d=mm&r=g)
I just released a new version (5.3) of MDT. Briefly, the changes relative to the last release are:
# Fixes to work with newer SWIG versions and the latest MODELLER API.
# A new function write_statpot() can generate a Modeller statistical potential file given a suitable input table.
# A new feature Cluster allows clustering of two features into a single one.
# A new feature AtomTable takes as input a table of precalculated per-atom values, and can be used to implement user-defined features or to use externally-calculated properties.
# To save space, the data for the MDT table itself can be compressed when writing to an HDF5 file with Table.write_hdf5().
# Certain library information (such as the mapping from feature values to bin indices, and atom or tuple class information) and information about the last scan is now written into MDT files in HDF5 format by Table.write_hdf5().
# The maximum value of bond_span_range can now be -1, to allow atom pairs that have no path of bonds between them. This is helpful to include inter-chain interactions, for example.
See http://salilab.org/mdt/ for download and documentation.
Ben
participants (1)
-
Ben Webb