Skip to main content

Table 10 Performance of a DN ensemble training on different groups of features on CASP9 targets

From: A study and benchmark of DNcon: a method for protein residue-residue contact prediction using deep networks

Feature set(s)

Acc. Top L/5 (SE)

Acc. Top L (SE)

 

Long

Medium

Long

Medium

seq

0.040(.005)

0.042(.004)

0.035(.003)

0.040(.003)

seq-atch

0.088(.006)

0.075(.006)

0.077(.005)

0.074(.003)

seq-atch-bins

0.078(.005)

0.092(.006)

0.077(.005)

0.085(.005)

seq-atch-bins-globals

0.142(.01)

0.202(.007)

0.100(.005)

0.130(.007)

seq-atch-bins-globs-pssm-ssa

0.157(.011)

0.221(.013)

0.106(.006)

0.132(.007)

pssm-atch

0.168(.012)

0.236(.014)

0.110(.006)

0.130(.007)

ALL

0.182(.012)

0.240(.013)

0.122(.006)

0.150(.007)