Skip to main content

Table 3 Average per-target correlation and loss for DeepQA and other top performing single-model QA methods on CASP11. The table is ranked based on the average per-target loss on stage two of CASP11. P-value of Wilcoxon signed ranked sum test* between DeepQA and other methods is also included in the table

From: DeepQA: improving the estimation of single protein model quality with deep belief networks

QA methods

Corr. on stage 1 /P-Value

Loss on stage 1 /P-Value

Corr. on stage 2 /P-Value

Loss on stage 2 /P-Value

DeepQA

0.64/-

0.09/-

0.42/-

0.06/-

ProQ2

0.64/4.80E-01

0.09/8.32E-01

0.37/2.84E-03

0.06/9.95E-01

Qprob

0.63/8.08E-01

0.10/9.38E-01

0.38/8.63E-03

0.07/7.12E-01

VoroMQA

0.56/1.60E-04

0.11/2.73E-01

0.40/2.57E-01

0.07/9.14E-01

ProQ2-refine

0.65/6.08E-02

0.09/9.17E-01

0.37/4.71E-03

0.07/4.86E-01

Wang_SVM

0.66/5.49E-02

0.11/7.98E-02

0.36/1.54E-02

0.09/4.91E-02

raghavagps-qaspro

0.35/3.79E-13

0.16/1.87E-04

0.22/1.92E-10

0.09/1.02E-03

Wang_deep_2

0.63/9.98E-01

0.12/7.18E-02

0.31/2.16E-06

0.09/8.22E-03

Wang_deep_1

0.61/3.06E-01

0.13/1.64E-03

0.30/5.93E-06

0.09/5.00E-03

Wang_deep_3

0.63/7.18E-02

0.12/3.15E-02

0.30/8.22E-03

0.09/8.22E-03

FUSION

0.10/8.43E-14

0.15/9.78E-04

0.05/1.81E-13

0.11/2.83E-07

RFMQA

0.54/1.61E-01

0.12/8.74E-01

0.29/3.80E-03

0.08/3.80E-03

ProQ3

0.65/1.62E-01

0.07/3.60E-02

0.38/4.44E-01

0.06/4.09E-01

ResQ*

0.67/-

0.05/-

0.58/-

0.09/-

ModFOLDclust2

0.74/3.96E-05

0.05/6.34E-04

0.56/1.80E-03

0.07/1.41E-01

Mean

0.57

0.11

0.33

0.08

  1. * The Wilcoxon signed ranked sum test was performed on the correlation and loss of targets between each method against DeepQA
  2. * ResQ was evaluated on 54 targets in CASP11, the local quality scores were converted into global quality score by equation \( Global=\frac{1}{L}{\sum}_{i=1}^L\frac{1}{1+{\left(\frac{Loca{l}_i}{5}\right)}^2} \). More detailed results can be found in Additional file 1: Table S4