Skip to content
Paul Lott edited this page Jan 23, 2013 · 7 revisions

The StochHMM model file contain the complete model topology (States, transitions) and values of the model. It can define which states call which user functions. However, user-functions must be defined before the executable is compiled.

The model divides sections (Model Information, States, Track Symbol Definitions) Divider lines shown as “====” or “####” are optional, but you’re encouraged to use them to make the model easier to read.

###5.3.1 Model Header (Required) The first line of a model file must contain the header.

#STOCHHMM MODEL FILE

###5.3.2 Model Information (Optional) Each model can contain a name, description, creation date information, creation commands, author, numerical attribute or range.

If a numerical attribute is assigned, a model function

#STOCHHMM MODEL FILE

MODEL INFORMATION
================================================
NAME:	Simplified Human Gene
DESCRIPTION:	Created using perl script
CREATION_DATE:	August 12,2009
CREATION_COMMAND:	./creation_script.pl fasta.fa....
AUTHOR:	Paul Lott
NUM_ATTRIB:	20
UPPER:	29
LOWER:	20

NAME => Name of the model
DESCRIPTION => Description of the model
CREATION_DATE => Date that the model was created
CREATION_COMMAND => Command or script used to create the model
AUTHOR => Who created the model
NUM_ATTRIB => Numerical value associated with the model. When using multiple models, StochHMM could select the model based on the value associated with the model. For example, if you have multiple models trained for sequences with different GC percentages. StochHMM could select the model based on the GC percentage of the sequence.
UPPER => Range value, similar to NUM_ATTRIB except defines the inclusive upper range of values associated with the model. If used LOWER must also be supplied.
LOWER => Range value, similar to NUM_ATTRIB except defines the inclusive lower range of values associated ith the model. If used UPPER must also be supplied.