FAQs
Table of Contents
-
What is SurvNet?
SurvNet is an application for identifying network-based biomarkers that most correlate with patient survival data. Using a biological network (e.g., a gene regulatory network or protein-protein interaction network) as the platform, SurvNet uses a systematic approach to identifying the subnetworks most associated with patient survival data.
The web app needs a network file (a human protein interaction network can be provided as default), a molecular profiling file (e.g., microarray gene expression, protein expression, DNA methylation or mutation data), and a patient survival data file as its input. The results will be displayed in a user-friendly way. SurvNet would be a valuable resource for biologists, clinicians, and bioinformaticians in the biomedical community.
-
Why SurvNet?
One major task in biomedical research is to identify the correlations between biomarkers and patient survival, which provides a foundation for diseases diagnostic and treatment. Conventionally, this kind analysis is performed based on individual genes. However, the results are often very noisy and hard to interpret at the mechanism level. Using a network as the platform, the network-based biomarkers are expected to boost the biological signals and return more robust results. More importantly, such biomarkers provide insights into the molecular mechanisms of diseases.
-
How does SurvNet work?
Network biomarkers have been a hot topic in the field and several studies have been published on this topic. Generally speaking, there are three components in an approach to identification of network-based biomarkers:
- network score function (summarizes node scores and network properties)
- searching algorithm
- statistical evaluation
The innovative aspect of SurvNet is to calculate node scores using the univariate Cox proportional hazards regression model, a widely used method for quantifying the correlations between gene activity and patient survival data (Cox and Oakes, 1984). For the searching algorithm and statistical evaluation, SurvNet uses the exact same schemes as described by Chuang et al., 2007 and Jia et al., 2011. Thus the underlying methodologies used by SurvNet are well-validated.
-
Does SurvNet require login?
No, login is not required — it is free for anyone to use.