Calling for Guidance on Interlayer Connection Analysis in Multilayer Networks

[Hui-SJTU]

When analyzing multilayer networks, how do you incorporate interlayer connections in the input data?
For example, in Structural Multiplex pipeline, the input folder contains two .xslx files, each for information about one layer, but how are the interlayer connections entered?
And how to calculate the multilayer measures that incorporate the interlayer connection information?
Thank you.

[c-yuwei]

Hi,

Thank you for reaching out to us, and for choosing to try BRAPH-2!

In the Structural Multiplex pipeline (braph2/pipelines/structural multiplex/example_ST_MP_WU.m), it employs a multiplex graph (using classes like MultiplexWU, MultiplexBUD, and MultiplexBUT), which is a type of multilayer graph where interlayer edges are only allowed between homologous nodes. In this case, the layers follow a categorical architecture, meaning all layers are interconnected. You can find more details http://braph.org/connectivity-analysis/multilayer-graphs/

Coming back to your question: “When analyzing multilayer networks, how do you incorporate interlayer connections in the input data?”
In BRAPH-2, interlayer connections are handled through multilayer graphs (classes like MultilayerWU, MultilayerBUD, and MultilayerBUT), which allow connections between any nodes across multiple layers.

This feature can be easily adapted within the current framework, and I’d be more than happy to assist. To make things smoother, we could arrange a brief meeting if that works for you.

Feel free to let me know, and thank you again for using BRAPH-2!

Best,
Yu-Wei

[Hui-SJTU]

Thanks, yuwei. I‘m very happy to get your reply so quickly. Regarding the first question, I would like to ask again, in the current pipeline of graph2, if I want to input a multiplex network, where the interlayer connections exist only on some of the nodes, as well as I want to give some weight to the interlayer connections, how can I do it?
And I would like to clarify the third question again, what multilayer metrics are currently computed in graph2 that incorporate information about interlayer connections?
Again, I am very much looking forward to the opportunity to have a meeting with you to discuss this issue further.

[c-yuwei]

Hi,

Thank you for your follow-up and clarification.

I will address your first and third questions directly one by one as follows:

Q1. In the current pipeline of graph2, if I want to input a multiplex network where the interlayer connections exist only on some of the nodes, and I want to assign weights to these interlayer connections, how can I do it?

In BRAPH-2, you can use classes such as MultilayerWU, MultilayerBUD, and MultilayerBUT. I can provide an example script after our discussion to better assist you.

Q3. What multilayer metrics are currently computed in graph2 that incorporate information about interlayer connections?

In BRAPH-2, both Multilayer Community (MultilayerCommunity) and Multilayer Modularity (MultilayerM) take interlayer connections into account. Following the naming pattern in BRAPH-2, under the folder /braph2/measures/, you can find all the graph measures included in the software. Measures with names containing Multiplex* are designed for multiplex graphs, while those with Multilayer* are for multilayer graphs (as explained in my previous reply). In this context, Multilayer Community (MultilayerCommunity) and Multilayer Modularity (MultilayerM) are currently designed for multilayer graphs.

BRAPH-2 also provides a highly flexible framework for implementing new graph measures. If you have specific graph measures or metrics you’d like to implement, it is feasible to do so quickly. These would seamlessly integrate with other BRAPH-2 functionalities, such as group comparisons, permutation tests, and brain visualisation.

Finally, regarding the meeting, please feel free to contact me at [email protected], and we can arrange a suitable time to discuss further.

Best,
Yu-Wei