DLP-SVM is a domain linker predictor.
It is composed of three loop-length dependent SVM predictors of domain
linkers (SVM-All, SVM-Long and SVM-Short),
and SVM-Joint, which combines
the results of SVM-Short and SVM-Long
into a single consolidated prediction (Ebina et al, 2009, Biopolymers ref 1).
The performances of our predictor appear to be largely related to the
quality of the domain data base, we developed in previous studies (ref 2 and 3).
An essential point of DLP is that only strictly-defined domain linkers were used to
train the predictor. Our definition of a domain linker is a loop sequence separating
two structural domains which can fold independently.
The training dataset is available here:
LinkerList.txt
Ebina T, Toh H, Kuroda Y:
Loop-length dependent SVM prediction of domain linkers for high-throughput structural proteomics. Biopolymers 2009;92(1):1-8.