
Note that only the lines starting with '#' are read by the program. 
All the other lines are just comments.

The mode is either 
DWT (full dinucleotide weight tensor), 
ADJ (a model with only dependencies between adjacent positions), or
PWM (the basic position specific weight matrix model)
#mode DWT

To control numerical stability of the determinant calculation, the
range of entries in the R matrix are limited to exp(precision). The
default value is 25.0
#precision 25.0

Whether to train a motif: 
0 = just calculates scores and outputs them, 
1 = train a motif model
#train_DWT 0

What type of background to use.
0 = uniform background probabilities, 
1 = set in the parameter file, or
2 = set from the input sequences
#background_type 1

Probabilities for each the four letters 
(only used when type is set to 1). 
#background_probs
A 0.2
C 0.3
G 0.3
T 0.2

Directory to write the output files.
If not specified, the default is the working directory. 
#output_dir Results

Minimal score (log-likelihood ratio under motif and backgroun models)
for printing out a site. The default is -50, which virtually
guarantees that a score is printed for every single sequence segment
in the input file. Note that this can lead to large output files.
#minimum_score 5.0


