Can spreading signals through a protein network find disease genes?
Spreading known disease-gene information smoothly across the whole protein interaction network ranked the true causal gene first more often than earlier network methods.
Source
Associating genes and protein complexes with disease via network propagation
Study at a glance
- Design
- Computational / modelling — Graph-based propagation of prior disease-similarity scores over a weighted human protein-protein interaction network, evaluated by leave-one-out cross-validation and on newly published gene-disease links.
- N
- N=1369 · Diseases in OMIM with a known causal gene used in cross-validation; separate validation used 51 new associations for 47 diseases plus 10 for diseases with previously unknown genes.
- Population
- OMIM disease-gene associations mapped onto a human protein-protein interaction network
- Outcome
- Rank of the hidden causal gene (top-1 success, precision-recall); coherence of inferred protein complexes
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
The authors built PRINCE, which gives each protein a prior score based on how similar its known diseases are to a query disease, then repeatedly passes those scores to network neighbours until they settle, so connected proteins end up with similar scores. They hid one known disease-gene link at a time and checked where PRINCE ranked the hidden gene within an artificial 100-gene interval, comparing against their reimplementations of a random-walk method and CIPHER. They also grew dense protein clusters from high-scoring proteins to propose disease-related complexes and examined three diseases in detail.
What they found
Across 1,369 OMIM diseases, PRINCE ranked the correct gene first in 34% of cases, versus 28.8% for random walk and 24.7% for CIPHER, and the ordering held for 2-, 5- and 10-fold cross-validation. On newly published associations it ranked a new gene first for 20 of 47 diseases. Its predicted complexes were more functionally coherent than a previous collection, and 61% had independent OMIM support compared with 7% for random gene sets.
The limits
What it doesn't show
PRINCE only works for diseases that resemble other diseases with known genes, because its signal comes from disease similarity. Its accuracy depends on how complete and accurate the protein interaction network is, which was far from complete in 2010. The candidate intervals were artificial 100-gene windows, not real linkage regions, and one leading competitor (Lage et al.) could not be reimplemented. Literature support for top predictions is partly circular, since many are well-known genes, and several mathematical symbols are missing from this text version, so the exact formulas cannot be checked here.
Key terms
- Network propagation
- Iteratively spreading scores from seed nodes to their neighbours in a graph until the values converge.
- Smoothness over a graph
- The assumption that connected nodes should have similar scores, used as a regulariser in semi-supervised learning.
- Gene prioritisation
- Ranking candidate genes by how likely they are to cause a given disease.
- Leave-one-out cross-validation
- Hiding one known example at a time, predicting it from the rest, and repeating for every example.
- Protein-protein interaction network
- A graph whose nodes are proteins and whose edges are experimentally observed physical interactions, here weighted by reliability.
Flashcards
0 of 10 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
What is the core idea of PRINCE's scoring?
Common questions
Why spread information globally rather than look only at direct neighbours?
Disease genes may be linked through longer paths; propagation lets evidence from distant but well-connected proteins contribute, which is what gave PRINCE its edge.
Why normalise edge weights by node degree?
Hub proteins interact with many others by chance, so down-weighting their edges stops them dominating every ranking.
Why hide all associations of the test protein, not just one?
To avoid easy cases where the same gene is linked to a very similar disease, which would inflate performance.
More on Graph neural networks