From: Improving protein secondary structure prediction based on short subsequences with local structure similarity
Word length n
6
7
8
9
Q3 (exact matching)
78.2
80.1
78.1
76.2
Q3 (inexact matching)
74.9
79.2
80.5
79.0