Skip to main content

Table 1 Probes correctly assigned with SLAHMMs that were not assigned with single-master HMMs, using a strict cutoff of 80 incorrect assignments (theoretical EPQ ~0.05).

From: Application of protein structure alignments to iterated hidden Markov model protocols for structure prediction

Probe SCOP ID

E-value

SCOP Superfamily

Incorrect Assignments

ND1: (EPQ = 1)

ND2: nr60 (EPQ ~0.05)

ND3: nr60 (EPQ = 1)

g1dy9.1

0.00018

b.47.1

0

 

X

 

d1k6wa2

0.00021

c.1.9

0

 

X

 

d1buoa_

0.00022

d.42.1

0

   

d1dm9a_

0.00039

d.66.1

0

 

X

 

d1rpxa_

0.00046

c.1.2

0

   

d1hq3d_

0.00067

a.22.1

0

   

d1at0__

0.00069

b.86.1

0

   

d1hq3b_

0.0011

a.22.1

0

   

d1kjqa2

0.0015

c.30.1

0

   

d7taa_1

0.0016

b.71.1

0

 

X

 

d2hrva_

0.0042

b.47.1

2

X

X

X

d1jfib_

0.0055

a.22.1

2

 

X

 

d1hq3a_

0.0057

a.22.1

2

   

d1bd0a2

0.0069

c.1.6

2

 

X

 

d1hx0a1

0.0096

b.71.1

2

 

X

 

d1bkra_

0.01

a.40.1

2

   

d1bd3a_

0.012

c.61.1

3

   

d1a0p_1

0.013

a.60.9

3

   

d1afra_

0.014

a.25.1

5

 

X

 

d1i6la_

0.016

c.26.1

6

X

X

X

d1k92a1

0.02

c.26.2

6

X

X

 

d1dfaa2

0.026

d.95.2

8

X

X

 

d1m4va1

0.031

b.40.2

8

 

X

 

d1gsoa2

0.036

c.30.1

8

X

X

X

d1j8ca_

0.04

d.15.1

8

X

  

d1dfca1

0.041

b.42.5

8

X

X

 

d1dxea_

0.041

c.1.12

8

 

X

 

d1efva2

0.048

c.31.1

9

 

X

 

d1k3sa_

0.063

d.198.1

13

 

X

 

d2pola2

0.066

d.131.1

14

X

X

 

d2pola3

0.071

d.131.1

15

   

d1k8kf_

0.1

d.198.2

24

X

X

X

d1efva1

0.11

c.29.1

24

X

X

 

d1j9qa2

0.13

b.6.1

29

   

d1gkpa1

0.14

b.92.1

36

   

d2a0b__

0.19

a.24.10

41

   

d1qo0d_

0.2

c.23.1

43

 

X

 

d1hava_

0.21

b.47.1

45

 

X

 

d1qg8a_

0.24

c.68.1

54

   

d1hq3c_

0.25

a.22.1

56

   

d1al3__

0.27

c.94.1

65

X

X

X

d1es9a_

0.28

c.23.10

68

   
  1. The "incorrect assignments" column refers to the number of incorrect assignments made at the point where the probe was correctly assigned a structure by a SLAHMM. The final three columns highlight the probes (marked with "X") that were not assigned by single-master HMMs even when they were allowed considerable advantages (the incorrect assignments column does not apply to these columns). ND1 ("Not Detected" 1) indicates the probes that could not be detected with a generous cutoff of up to 1575 incorrect assignments (theoretical EPQ = 1). ND2 and ND3 present results where the single-master HMMs were allowed to search the entire nr60 database for additional homologs during the training phase (see text for details).