Title: Biological Inferences Using Networks Measures Abstract: ======== Networks, called graphs by mathematicians, have been used to study many areas of biology. In many systems, the connections between elements (such as genes, proteins, or diseases) are at least as important as the elements themselves, and network theory gives a framework for studying these connections. Network measures can be used to generate hypotheses about biological data that can be tested in a wet lab. Here, I present three present three of the network methods I have developed to gain insights from biological data. First I will show how a metric from graph theory, k-connectivity, which is seldom used in biology, can be used to find biologically significant modules. Next, I will discuss how differentially expressed gene sets appear in regulatory networks and how taking this network perspective into account can give a better null model for determining which gene-classes significantly overlap with these sets. Finally, I will discuss a model for biological data that uses hypergraphs, an extension of graphs, and the development of metrics to assist in the analysis of this data. Many different types of biological data currently modeled with networks can be better modeled with hypergraphs, but there are challenges associated with the hypergraph model due to the fact that many of the metrics used on networks do not easily translate to the hypergraph realm. I will discuss two hypergraph metrics that I have developed and show examples of where these allows for better predictions than a simple network model would.