3D-Jury is a metaserver that aggregates and compares models from various protein structure prediction servers.[1]

screenshot of the 3D-Jury minimalist web interface, showing three text boxes in the center, text near the bottom with some server metadata, and the right side with check boxes of different servers to use in the algorithm
Screenshot of Structure Prediction Meta Server, 3D-Jury, web interface from the Wayback Machine

The 3D-Jury algorithm takes in groups of predictions made by a collection of servers and assigns each pair a 3D-Jury score, based on structural similarity. To improve accuracy of the final model, users can select the prediction servers from which to aggregate results.[1] The authors of 3D-Jury designed the system as a meta-predictor because earlier results concluded that the average low-energy protein conformation (by way of aggregation) fit the true conformation better than simply the lowest-energy protein conformation.[2]

The Robetta automatic protein structure prediction server incorporates 3D-Jury into its prediction pipeline.[3]

As of January 2024, the links to 3D-Jury originally hosted by the BioInfoBank Institute are no longer valid.[4]

Algorithm

edit

First, pairwise comparisons are made between every combination of models generated from chosen protein prediction servers. Each comparison is then scored using the MaxSub tool.[5] The score,  , is generated by counting the number of Cα atoms in the two predictions within 3.5 Å of each other after being superpositioned.

To get a roughly 90% chance two models are of a similar fold class, the authors set a threshold of 40 as the lowest score possible for a pair of models to be annotated as "similar".[1] The authors admittedly chose this threshold based on unpublished work.

There are two scores 3D-Jury gives: the best-model-mode score using one model from each server ( ) and the all-model-mode score that considers all models from each server ( ).[1]

The best-model-mode score using one model per server,  , is calculated as,

 

where   is the number of servers and   is the number of top ranking models (with a maximum of 10) from the server  , while a pairwise similarity score is calculated between models   (model   from server  ) and   (model   from server  ).[1]

While the all-model-mode score considering all models from the servers,  , is calculated as,

 

using similar variables as noted with the best-model-mode score.

Note, these meta-predictor scores do not take into account the confidence scores from each of the models from other servers.[1]

References

edit
  1. ^ a b c d e f Ginalski K; et al. (2003). "3D-Jury: a simple approach to improve protein structure predictions". Bioinformatics. 19 (8): 1015–1018. doi:10.1093/bioinformatics/btg124. ISSN 1367-4803. OCLC 110817016. PMID 12761065.
  2. ^ Bonneau, Richard; Ruczinski, Ingo; Tsai, Jerry; Baker, David (2002). "Contact order and ab initio protein structure prediction". Protein Science. 11 (8): 1937–1944. doi:10.1110/ps.3790102. ISSN 0961-8368. OCLC 112117834. PMC 2373674. PMID 12142448.
  3. ^ Chivian D; et al. (2005). "Prediction of CASP6 structures using automated Robetta protocols". Proteins. 61 (S7): 157–166. doi:10.1002/prot.20733. PMID 16187358. S2CID 8122486. Archived from the original on 2012-12-10.
  4. ^ "BioInfoBank Meta Server". BioInfoBank Meta Server. Archived from the original on 2007-01-13. Retrieved 2024-01-17.
  5. ^ Siew, Naomi; Elofsson, Arne; Rychlewski, Leszek; Fischer, Daniel (2000-09-01). "MaxSub: an automated measure for the assessment of protein structure prediction quality". Bioinformatics. 16 (9): 776–785. doi:10.1093/bioinformatics/16.9.776. ISSN 1367-4803. OCLC 121793099. PMID 11108700.
edit