logo Netzschleuder network catalogue, repository, and centrifuge

human_symptoms_diseases — Human symptoms-disease network (2014)

Description

Networks of human diseases and their symptoms, extracted from the biomedical literature by Zhou et al. The authors extracted the 7,109,429 records of the PubMed bibliographic database with at least one disease or symptom term (according to the Medical Subject Headings, MeSH) in their MeSH metadata field, yielding 4,442 disease terms and 322 symptom terms; 849,103 records contain at least one disease and one symptom term.

  • symptoms_diseases: bipartite network of symptoms and diseases, where an edge connects a symptom and a disease that co-occur in the MeSH metadata of at least one PubMed record (322 symptoms and 4,219 diseases). The edges are weighted by the number of co-occurrences, and by their term frequency-inverse document frequency (TF-IDF), which discounts symptoms that occur with many diseases: w = W log(N / n), where W is the number of co-occurrences, N the number of diseases and n the number of diseases in which the symptom appears. From these weights, the authors computed the symptom similarity between diseases as the cosine similarity of their vectors of symptom weights, which defines the (dense) human symptoms disease network (HSDN).
  • disease_links: network of the 133,106 links of the HSDN between 1,596 diseases that have both significant shared symptoms and shared genes or protein-protein interactions (as reported in Supplementary Data 4 of the paper), weighted by their symptom similarity.

The networks have the following properties:

  • Nodes: name, the MeSH term; pubmed_occurrence, the number of PubMed records with the term in their MeSH metadata.
  • symptoms_diseases nodes: is_symptom, whether the node is a symptom (otherwise a disease). 248 terms are listed both as diseases and as symptoms in the source data, and correspond to two different nodes.
  • symptoms_diseases edges: occurrence, the number of PubMed records in which the symptom and disease co-occur; tfidf, the TF-IDF weight of the association.
  • disease_links edges: weight, the symptom similarity of the two diseases.
Tags
Biological Public Health Weighted Metadata
Citation
  • X. Zhou, J. Menche, A.-L. Barabási, A. Sharma, "Human symptoms-disease network." Nature Communications 5, 4212 (2014) 10.1038/ncomms5212 Sci-Hub
Upstream URL OK
https://doi.org/10.1038/ncomms5212
Networks
Tip: click on a column header to sort the list, or hover over it for a description.
Column legend
$\left<k\right>$
Mean degree
$\sigma_k$
Standard deviation of the degrees
$\lambda_h$
Largest eigenvalue of the Hashimoto (non-backtracking) matrix
$\tau$
Random walk mixing time
$r$
Degree assortativity
$c$
Global clustering coefficient
$\oslash$
(Pseudo-) diameter
$S$
Relative size of the largest component
NPs
Node properties
EPs
Edge properties
Name Nodes Edges $\left<k\right>$ $\sigma_k$ $\lambda_h$ $\tau$ $r$ $c$ $\oslash$ $S$ Kind Mode NPs EPs gt GraphML GML csv
symptoms_diseases 4,541 147,978 65.17 166.76 234.98 2.90 -0.52 0.00 6 1.00 Undirected Bipartite name is_symptom pubmed_occurrence occurrence tfidf 401 KiB 944 KiB 914 KiB 707 KiB