Analysis of Genes Involved in Human Cancer
Author Information
Author(s): Furney Simon J, Higgins Desmond G, Ouzounis Christos A, López-Bigas Núria
Hypothesis
What are the unique features of genes involved in cancer that can assist in identifying new candidate cancer genes?
Conclusion
The study developed a Bayesian classification model that scores human genes for their likelihood of involvement in cancer based on their structural and functional properties.
Supporting Evidence
- The study identified 291 genes known to be involved in cancer.
- Cancer genes show higher conservation scores compared to the rest of the human proteome.
- The Bayesian model predicted 2295 human genes likely to be involved in cancer.
Takeaway
Scientists looked at genes that cause cancer to find patterns that could help identify more cancer-related genes. They created a model to predict which genes might be involved in cancer.
Methodology
The study analyzed gene properties and developed a Bayesian classifier based on sequence properties and Gene Ontology terms.
Potential Biases
Some genes selected as negative examples may actually be involved in cancer, leading to potential bias in predictions.
Limitations
The model may include false positives as it is difficult to definitively identify genes not involved in cancer.
Statistical Information
P-Value
1.57e-09
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website