Plant protein-protein interaction sites

1   What is PlaPPISite?

PlaPPISite is a database that presents fine-grained interaction details for 13 plant PPIs, including 121 PPIs with experimental complex structures and 132,328 PPIs whose interaction details are inferred from two distinct computational pipelines as follows.

First, the 3D structures of 132,328 PPIs are modeled by using Homology Modeling of Protein Complex (HMPC) technique or the Protein Interactions by Structural Matching (PRISM) software. Multiple complex structures based on different template complexes could be predicted for a PPI, however, only the one with the highest quality considering the sequence identity, coverage and the resolution is displayed in PlaPPISite. For each predicted complex structure, PlaPPISite not only provides an interactive user interface for browsing the interaction sites in the context of 3D structures, but also lists detailed evolutionary and physicochemical properties (e.g. residue conservation, and estimated energy score) of these sites (Example).

Second, PlaPPISite assigns the domain-domain interactions or the domain-motif interactions to 346,930 PPIs whose 3D structures cannot be modeled. In this case, the users can easily query the protein interaction regions at the sequence level (Example).

PlaPPISite is a free and user-friendly database. It does not require user registration or any configuration on local machines.

PPI distribution:



2   Flow diagram of database construction

3  Detailed protein interaction information

3.1   Protein alignment information

For each PPI with available structure, the alignment information between target and template has been showed as sequences format in PlaPPISite. Besides, the identity and coverage of the alignment was also provided.

An example of alignment information for the single protein of predicted 3D complex structure:

3.2   Protein interaction networks

There are 4 kinds of PPI networks in PlaPPISite.

(1)  PPI network of experimentally verified PPIs (HMPC/PRISM).

(2)  PPI network of experimentally verified PPIs (DDI/DMI).

(3)  PPI network of predicted PPIs (HMPC/PRISM).

(4)  PPI network of predicted PPIs (DDI/DMI).


An example of searching experimentally verified PPI network:



An example of searching predicted PPI netwoek:



3.3   Protein complex structure and interaction site information

Choose interested residue pairs and click the show button, then the residue pairs will be highlighted on the 3D structure.

What's more, click the spin button, then the structure will be rotated with multi-angle.





3.4   DDI/DMI annotation information

For PPIs whose structure were unresolved, users can retrieve the corresponding DDI or DMI annotation.

3.5   Annotation information of the predicted PPIs

Predicted PPIs were annotated from several aspects to evaluate the reliability of the prediction.Such as interolog species,GO annotations, subcellular localizations, expression data of predicted PPI and corresponding similarity information etc.

4   Key word search(single protein & protein pairs)

Users can search single protein or protein pairs by Uniprot ID, GeneID, Gene name and protein name etc. The web provides multiple examples about protein search and PPI search.


5   The online prediction platform of complex structure

PlaPPISite provides the online prediction platform for users to predict their interested complex structure by using HMPC and PRISM methods.Users only need to submit the sequences of a pair of PPI and choose the interested prediction method(HMPC or PRISM).What's more, to save time, we request users fill in the email address. As the process of prediction was done, we will send the email information of the prediction result to the user.

On the result page, users can download the complex structure, single protein structure, predicted interaction sites and initial protein sequences etc.

6   How to evaluate interested 3D structures of protein complexes?

To evaluate the quality of HMPC and PRISM complex structures in PlaPPISite, 121 PPIs with experimental complex structures are served as a test set (i.e., golden standard), in which 74 and 23 PPIs can be modeled using HMPC and PRISM respectively.

According to the evaluation criterion of Critical Assessment of Predicted Interactions (CAPRI), which is a community-wide experiment to assess the accuracy of predicted 3D structures of protein complexes, the predicted protein complex structures can be grouped into four categories on the basis of the backbone root-mean-square deviation of interface residues (I_RMS): high accuracy (I_RMS ≤ 1.0Å), medium accuracy (1.0Å < I_RMS ≤ 2.0Å), acceptable (2.0Å < I_RMS ≤ 4.0Å) and incorrect (I_RMS > 4Å). With respect to the 74 complex structures predicted by HMPC, 4 (5.4%) complex structures achieve high accuracy, 74 (100.0%) complex structures are predicted correctly, and 0 (0.0%) complex structures are predicted incorrectly (Table 1). Regarding the 23 complex structures predicted by PRISM, 6 (26.1%) complex structures achieve high accuracy, 22 (95.7%) complex structures are predicted correctly, and 1 (4.3%) complex structures are predicted incorrectly (See the following table). Comparatively, the accuracy of PRISM is inferior to that of HMPC, meaning that more cautions should be taken when dealing with the complex structures as well as the corresponding interaction sites inferred from PRISM.





7   How to interact with 3D structures by mouse operations?

Visualization of protein structures is implemented by NGL, which is a WebGL based 3D viewer powered by MMTF.

Movement    Mouse Input  Touch Input
RotationPrimary Mouse ButtonSingle Touch
TranslationMiddle Mouse Button or Ctrl+PrimaryTriple Touch
ZoomScroll Wheel or Second Mouse Button or Shift+PrimaryPinch (double touch)
SlabCtrl+SecondNot Available

8   How to download 3D structures of protein complexes?

Users can download the pdb file of interested protein structure by clicking the download button in the result page. Alternatively, PlaPPISite allows the users to download all of the data for further analyses in the download page.

9  How to cite?

Li H, Yang S, Wang C, Zhou Y, Zhang Z. AraPPISite: a database of fine-grained protein–protein interaction site annotations for Arabidopsis thaliana. Plant Mol Biol. 2016;92:105–116. Link
Yang X, Li H, Yang S, Qi H, Wang T, Zhang Z. PlaPPISite: a comprehensive resource for plant protein-protein interaction sites. BMC Plant Biol. 2020;20:61. Link



Copyright © 2021 Ziding Zhang's Lab - China Agricultural University. All Rights Reserved. Maintained by Xiaodi Yang   京ICP备18035355号