Question

This question refers to the inference of association networks, as presented in this Lesson. Which of the following statements is/are TRUE?

O Negative correlations imply that the nodes should not be connected.
O The Fisher-transformed correlation value of two random vectors follows a Gaussian distribution with zero mean and variance 1/(n-3), where n is the dimension of each random vector.
O The Multiple Testing problem is harder in network inference than in other contexts because the involved statistical tests are typically not independent.
O The association network inference framework, as presented in this Lesson, can also be applied in directed networks.

Answer 1

In the inference of association networks, negative correlations simply indicate an inverse relationship without preventing connections, and Fisher-transformed correlation values follow a Gaussian distribution.

Step-by-step explanation:

Regarding the inference of association networks, it is true that negative correlations do not necessarily imply that nodes should not be connected; in fact, they simply indicate an inverse relationship between the nodes. It is also true that the Fisher-transformed correlation value of two random vectors tends to follow a Gaussian distribution with a mean of zero and variance of 1/(n-3), with 'n' being the dimension of each random vector.

Additionally, the Multiple Testing problem is indeed more challenging in network inference compared to other contexts because the statistical tests are often not independent. Lastly, while the association network inference framework described in the lesson could potentially be adapted to directed networks, it is not inherently designed for them and may require additional considerations. Multiple Testing is more challenging in network inference, and the framework can potentially be adapted to directed networks.