Applying Monte Carlo Simulation to Biomedical Literature to Approximate Genetic Network

Rami Al-Dalky, Kamal Taha, Dirar Al Homouz, Murad Qasaimeh

Research output: Contribution to journalArticlepeer-review

15 Scopus citations


Biologists often need to know the set of genes associated with a given set of genes or a given disease. We propose in this paper a classifier system called Monte Carlo for Genetic Network (MCforGN) that can construct genetic networks, identify functionally related genes, and predict gene-disease associations. MCforGN identifies functionally related genes based on their co-occurrences in the abstracts of biomedical literature. For a given gene g , the system first extracts the set of genes found within the abstracts of biomedical literature associated with g. It then ranks these genes to determine the ones with high co-occurrences with g. It overcomes the limitations of current approaches that employ analytical deterministic algorithms by applying Monte Carlo Simulation to approximate genetic networks. It does so by conducting repeated random sampling to obtain numerical results and to optimize these results. Moreover, it analyzes results to obtain the probabilities of different genes' co-occurrences using series of statistical tests. MCforGN can detect gene-disease associations by employing a combination of centrality measures (to identify the central genes in disease-specific genetic networks) and Monte Carlo Simulation. MCforGN aims at enhancing state-of-the-art biological text mining by applying novel extraction techniques. We evaluated MCforGN by comparing it experimentally with nine approaches. Results showed marked improvement.

Original languageBritish English
Article number7274679
Pages (from-to)494-504
Number of pages11
JournalIEEE/ACM Transactions on Computational Biology and Bioinformatics
Issue number3
StatePublished - 2016


  • biological NLP
  • biomedical literature
  • Gene Regulatory Network
  • gene-disease associations
  • information extraction
  • Monte Carlo simulation
  • Text mining


Dive into the research topics of 'Applying Monte Carlo Simulation to Biomedical Literature to Approximate Genetic Network'. Together they form a unique fingerprint.

Cite this