[ Start ]

The basis behind molecular modeling is to use as much information as possible derived from solved structures in the PDB and apply them to the wealth of newly generated gene sequences, derived from many genome programs. All the available parameters are considered. Whenever there are variables that are too uncertain to derive from experimental data, you can use powerful prediction algorithms such as the ICM program to find the most probable solution. With today's need for high-throughput, molecular modeling is often one of the best approaches to define priorities for researchers and corporations.

ICM has an excellent record in building accurate models by homology. The procedure will build the framework and shake up the side-chains and loops by global energy optimization. You can also color the model by local reliability to identify potential errors in your model.

ICM also offers a fast and completely automated method to build a model by homology and extract the best fitting loops from a database of all known loops. It just takes a few seconds to build a complete model by homology with loops.

20.1.1 Getting Started

The three items you need for ICM protein molecular modeling are:

1. An alignment (see alignment section) of your protein sequence against a template structure from the PDB. This is a *.ali file in ICM.
2. A template structure from the PDB converted (See convert object) into an ICM object.
3. A sequence file of your query sequence for the structure you wish to allographic-construct and a sequence file for your template. Note ICM automatically extracts this information from the alignment or template structure.

Your graphical user interface window should look something like this: