spacer

CAPRI: Critical Assessment of PRediction of Interactions

Community wide experiment on the comparative evaluation of protein-protein docking for structure prediction

Hosted By EMBL/EBI-PDBe Group

Round 28

Round 28 ID mapping from group number to Accessor Number for Target 60
Round 28 ID mapping from group number to Accessor Number for Target 61
Round 28 ID mapping from group number to Accessor Number for Target 62
Round 28 ID mapping from group number to Accessor Number for Target 63
Round 28 ID mapping from group number to Accessor Number for Target 64


Round 28 ID mapping from group number to Accessor Number for Target 59
Round 28 UPLOADER ID mapping for Target 59
Round 28 SCORER ID mapping for Target 59

NOTE The Coordinates are not yet available for download

T60 result summary
T61 result summary
T62 result summary
T63 result summary
T64 result summary

T59 result summary

Short description of the peptide assessment procedure

With respect to the regular assessment of protein-protein interaction
in CAPRI, the assessment criteria have been slightly tightened for the
assessment of protein-peptide interaction.
The following distance cut-offs are employed:

  • a distance cut-off of 8 A between any two CB atoms (or CA for Gly) to define interface residues
  • a distance cut-off of 4 A between any two atoms to define native contacts
  • a clash distance of 3 A
Also the classification criteria have been tightened:
  • f(nat) thresholds of 0.2, 0.5 and 0.8 now distinguish between high (f(nat)>0.8), medium (f(nat)>0.5), acceptable (f(nat)>0.2) and incorrect models (f(nat)<0.2), or:
    * fnat: incorrect < 0.2 < acceptable < 0.5 < medium < 0.8 < high
  • and similarly L-rms and I-rms:
    * lrms: incorrect > 4.0 > acceptable > 2.0 > medium > 1.0 > high
    * irms: incorrect > 2.0 > acceptable > 1.0 > medium > 0.5 > high
in other words: a high-quality protein-peptide model has f(nat) above 0.8 and either L-rms below 1.0 or I-rms below 0.5.

Finally, a sequence overlap of 90% with the target is required for peptides.

Target 60

Results T60 interfaces A:B - click here to see the results
Results T60 A:B interface is taken, but the peptide is shortened to correspond to the same residues as A:C- click here to see the results
Results T60 interfaces A:C- click here to see the results
Full results:
Clash threshold25.74
average12.32
std dev6.71


T60.ABT60.ACT60.AB_6
Nr groups 32 32 32
High Accuracy (***) 0 ( 0) 25 ( 7) 3 ( 2)
Medium (**) 1 ( 1) 9 ( 5) 59 (12)
Acceptable (*) 3 ( 2) 22 ( 8) 25 (10)
Incorrect 301 (32) 251 (31) 220 (26)
Clashes 13 ( 7) 13 ( 7) 13 ( 7)
Low ID 2 ( 1) 0 ( 0) 0 ( 0)
Total 320 (32) 320 (32) 320 (32)


Target 61

Results T61 interfaces A:B - click here to see the results
Results T61 A:B interface is taken, but the peptide is shortened to correspond to the same residues as A:C- click here to see the results
Results T61 interfaces A:C- click here to see the results
Full results:
Clash threshold30.98
average13.59
std dev8.69


T61.ABT61.ACT61.AB_6
Nr groups 32 32 32
High Accuracy (***) 0 ( 0) 13 ( 3) 7 ( 1)
Medium (**) 1 ( 1) 17 ( 7) 45 (11)
Acceptable (*) 4 ( 3) 22 (10) 32 (12)
Incorrect 299 (32) 254 (31) 222 (27)
Clashes 14 ( 4) 14 ( 4) 14 ( 4)
Low ID 2 ( 1) 0 ( 0) 0 ( 0)
Total 320 (32) 320 (32) 320 (32)


Target 62

Results T62 interfaces A:B - click here to see the results
Results T62 A:B interface is taken, but the peptide is shortened to correspond to the same residues as A:C- click here to see the results
Results T62 interfaces A:C- click here to see the results
Full results:
Clash threshold40.46
average13.96
std dev13.25


T62.ABT62.ACT62.AB_6
Nr groups 32 32 32
High Accuracy (***) 0 ( 0) 28 ( 7) 5 ( 1)
Medium (**) 1 ( 1) 6 ( 4) 58 (11)
Acceptable (*) 5 ( 4) 16 ( 9) 26 ( 9)
Incorrect 293 (31) 251 (30) 212 (26)
Clashes 9 ( 3) 9 ( 3) 9 ( 3)
Low ID 12 ( 2) 0 ( 0) 0 ( 0)
Total 320 (32) 310 (31) 310 (31)


Target 63

Results T63 interfaces A:B - click here to see the results
Results T63 A:B interface is taken, but the peptide is shortened to correspond to the same residues as A:C- click here to see the results
Results T63 interfaces A:C- click here to see the results
Full results:
Clash threshold30.88
average13.38
std dev8.75


T63.ABT63.ACT63.AC_6
Nr groups 32 32 32
High Accuracy (***) 25 ( 8) 0 ( 0) 0 ( 0)
Medium (**) 10 ( 5) 2 ( 1) 10 ( 5)
Acceptable (*) 30 (10) 16 ( 8) 60 (15)
Incorrect 244 (30) 293 (32) 239 (29)
Clashes 7 ( 2) 7 ( 2) 7 ( 2)
Low ID 3 ( 1) 1 ( 1) 3 ( 1)
Total 319 (32) 319 (32) 319 (32)


Target 64

Results T64 interfaces A:B - click here to see the results
Results T64 A:B interface is taken, but the peptide is shortened to correspond to the same residues as A:C- click here to see the results
Results T64 interfaces A:C- click here to see the results
Full results:
Clash threshold31.61
average14.17
std dev8.72


T64.ABT64.ACT64.AB_6
Nr groups 32 32 32
High Accuracy (***) 0 ( 0) 21 ( 7) 0 ( 0)
Medium (**) 1 ( 1) 11 ( 5) 54 (10)
Acceptable (*) 11 ( 4) 14 ( 8) 26 (11)
Incorrect 295 (32) 260 (31) 226 (28)
Clashes 10 ( 3) 10 ( 3) 10 ( 3)
Low ID 1 ( 1) 2 ( 2) 2 ( 2)
Total 318 (32) 318 (32) 318 (32)


Target 59

Results T59 - click here to see the results

Full results (regular assessment):
Clash threshold35.62
average10.63
std dev12.50


PredictorUploaderScorer
Nr groups401724
High Accuracy (***)0 (0)0 (0)0 (0 <- 0)
Medium (**)1 (1)1 (1)0 (0 <- 0)
Acceptable (*)21 (12)23 (9)9 (8 <- 4)
Incorrect282 ( 32)1320 (14)186 (23 <- 17)
Clashes16 ( 6)55 (5)25 (5 <- 8)
Low Id49 ( 5)283 (3)15 (6 <- 3)
Total369 (37)1682 (17)235 (24 <- 17)

Numbers in parentheses refer to the number of different source groups producing these models
(ie z scorer groups recognize y out of x uploader groups)


ROUND 28 Timeline

Round 28 will have following targets :
Target 59 : A protein-protein complex
Target 60-64 : protein-peptide complexes

  • 8th April 2013 - Registration opens for Round 28
  • 24th April 2013 - Round 28 opens - Prediction of target 59 (a protein-protein complex) and 60-64 (protein-peptide complexes)
  • 5th May 2013 - Deadline for submitting models for target 59 (a protein-protein complex)
  • 12th May 2013 - Deadline for submitting models for targets 60-64 (protein-peptide complexes)
  • 8th May 2013 - Scoring starts for target 59 scoring server
  • 12th May 2013 - Deadline for scoring submission for target 59 scoring server


Documentation and Targets Round 28

  • Participants
  • Management Group
  • Prediction formats
  • Download Opens 24th April 2013
  • Upload Opens 24th April 2013

  • Entries

  • Current submissions

  • Email Problems or Queries to Sameer Velankar

    spacer
    spacer