DIRseq is a sequence-based method for predicting drug-interacting residues of an intrinsically disordered protein. For a central residue n, every other residue i contributes a multiplicative factor f(i; n), which depends on the amino-acid type of residue i and the sequence distance |i-n|. The total factor of residue n is then converted to a propensity score via a sigmoid function.
Name:
Seqence:
1st column, amino acid; 2nd column, predicted R2 (in s-1)
[Note: a uniform scaling factor may be required to get best match with measured results.]