PROPER: a human PPI reference network generated by PROER-seq technology

PROPER v.1.0

PROPER v.1.0 is a human PPI reference network generated by PROER-seq technology.

The understanding of protein-protein interactions (PPIs) has expanded our abilities to interpret the genome. We developed the PROPER-seq (protein-protein interaction sequencing) technology to map PPIs at the transcriptomic scale. Starting from cells, a typical laboratory can map 10,000 to 100,000 PPIs within a few weeks. PROPER-seq converts the input transcriptome into a library of RNA-barcoded proteins, and subsequently reverse transcribes and ligates the two RNA barcodes of each PPI to produce a chimeric DNA sequence. This approach enables simultaneous conversion of many PPIs into chimeric DNA sequences and subsequent decoding of the PPIs by sequencing these chimeric sequences. Applied to human embryonic kidney, T lymphocyte, and endothelial cells, PROPER-seq revealed 210,518 PPIs (collectively called PROPER v.1.0). PROPER v.1.0 exhibited higher validation rates than the PPIs identified by other high-throughput methods when compared to prior literature. Four out of the 4 novel PPIs that we tested were validated by in vivo experiments. PROPER-seq offers orders of magnitude gain in throughput for PPI mapping and PROPER v1.0 provides a reference human PPI network.


Query network visualization

Node information table

Edge information table

Node information table

Edge information table

Download the Full PROPER v.1.0 Data

Download link

Description of the format

The full PROPER v1.0 data file is a simple table of all the edges in csv format. There are 7 columns:

  • Gene1: Gene symbol
  • Gene2: Gene symbol
  • Cell line specificity: HEK, Jurkat, HUVEC, or shared
  • Read count: Number of read pairs mapped in the merged PROPER-seq library
  • Odds ratio: Odds ratio of the PPI derived from the chi-square test of calling PPIs from the merged PROPER-seq library
  • BH-corrected p-value: Benjamini-Hochberg corrected p-values derived from the chi-square test of calling PPIs from the merged PROPER-seq library
  • Potential background contamination: A binary indication of whether the PPI may be resulted from potential background contaminations.

Statistics of the data

  • Total nodes (genes): 8,635
  • Total edges (interactions): 210,518


1. The PROPER-seq technology and data processing methods

Our manuscript of PROPER-seq is in preparation.

  • Kara L. Johnson, Zhijie Qi, Zhangming Yan, Xingzhao Wen, Tri C. Nyugen, Kathia Zaleta-Rivera, Chien-Ju Chen, Xiaochen Fan, Kiran Sriram, Xueyi Wan, Zhen Bouman Chen, Sheng Zhong. Revealing protein-protein interactions at the transcriptomic scale by sequencing

2. How to use the PROPER v.1.0 database?

The PROPER v.1.0 database is a user friendly web app for querying and downloading data of the human reference PPI network generated by PROPER-seq technology in three cell lines, which are HEK, HUVEC and Jurkat cell lines.

Threr are two data query approaches:

  • Vis-Search: query genes with network visualization.
  • Table-Search: query genes in data tables.

The Vis-Search approch is limited by the size of query sub-network. When the query result is a network with more than 10,000 edges, the Vis-Search approach will be terminated. Please input less query genes or use Table-Search approach.

The PROPER v.1.0 database is still under construction. We will improve its functionality.