This directory contains the data used in the experiments in 
"Classifying GPCRs with Support Vector Machines"

The sequences and subfamilies were taken from GPCRDB in September of 2000, 
and no longer are an accurate reflection of the  composition or organization 
of GPCRDB.

----------------------------------------------------------------------------


Each sequence file contains positive and negative examples for training a support vector machine to recognize the subfamily. 

For our cross-validation protocol, the entire dataset of sequences was randomly partitioned into set0 and set1.

For each set0 subfamily sequence file, the positive examples are the members of that subfamily in set0 and the negative examples are the remaining sequences in set0 (and likewise for set1).

Formatting:

All positive examples are denoted with a "1"

>1_O08620|O08620
MLLLLLVPLF LRPLGAGGAQ . . .

All negative examples are denoted with a "0"

>0_O02464|O02464
MDPGPGLAAL QAWAAKSP . . .