The following sections briefly describes DAMAVER, how to run DAMAVER from the command
line, as well as the required input and the produced output files. Previously, DAMAVER
used to have a set of companion applications, DAMSEL, DAMSUP, DAMFILT, DAMSTART and
DAMCLUST. The functionality of all these have been merged into DAMAVER for ease of use.
If you use results from DAMAVER in your publication, please cite:
DAMAVER is an application to (a) align ab initio low resolution models
e.g. provided by DAMMIF, DAMMIN
or GASBOR; (b) select the most typical ("probable")
model; and (c) and provide averaged, filtered and model suitable as search volume
input for DAMMIN (a.k.a. DAMSTART model).
Given several ab initio models in PDB or mmCIF format, the program first
superimposes all possible pairs by applying one of the methods for superposition of
models that are available in CIFSUP, namely NSD, NCC or ICP,
without need of calling externals programs. The most representative model is determined,
and outlier are excluded from subsequent calculations. The aligned models are remapped
onto a grid of densely packed beads in order to compute a frequency map.
For each bead, the cross volume with proximal dummy atoms in the input models (occupancy)
is computed and saved into the output file (damaver model).
This model is also filtered, i.e. unlikely beads removed (damfilt model), and
surrounded with a solvent shell to make it suitable as a start model for
DAMMIN (damstart model).
It is to note that none of the models generated, damaver, damfilt or damstart
will generally fit the experimental data. Use the damstart model with
DAMMIN to obtain a final model that fits the data.