Rxivist logo

A reliable and unbiased human protein network with the disparity filter

By Gregorio Alanis-Lobato, Miguel A. Andrade-Navarro

Posted 23 Oct 2017
bioRxiv DOI: 10.1101/207761

The living cell operates thanks to an intricate network of protein interactions. Proteins activate, transport, degrade, stabilise and participate in the production of other proteins. As a result, a reliable and systematically generated protein wiring diagram is crucial for a deeper understanding of cellular functions. Unfortunately, current human protein networks are noisy and incomplete. Also, they suffer from both study and technical biases: heavily studied proteins (e.g. those of pharmaceutical interest) are known to be involved in more interactions than proteins described in only a few publications. Here, we use the experimental evidence supporting the interaction between proteins, in conjunction with the so-called disparity filter, to construct a reliable and unbiased proteome-scale human interactome. The application of a global filter, i.e. only considering interactions with multiple pieces of evidence, would result in an excessively pruned network. In contrast, the disparity filter preserves interactions supported by a statistically significant number of studies and does not overlook small-scale protein associations. The resulting disparity-filtered protein network covers 67% of the human proteome and retains most of the network's weight and connectivity properties.

Download data

  • Downloaded 307 times
  • Download rankings, all-time:
    • Site-wide: 55,405 out of 94,912
    • In bioinformatics: 6,132 out of 8,837
  • Year to date:
    • Site-wide: 90,195 out of 94,912
  • Since beginning of last month:
    • Site-wide: 75,676 out of 94,912

Altmetric data

Downloads over time

Distribution of downloads per paper, site-wide


Sign up for the Rxivist weekly newsletter! (Click here for more details.)