- This topic has 3 replies, 2 voices, and was last updated 3 years ago by
.
-
Posts
-
On fMRI multiplex networks:
1) I have calculated my own supra-adjacency matrices. Is there a way to run the multiplex pipeline using exclusively those?
2) It is not clear to me how the multiplex matrices are constructed using the example data given. For the “connectivity-functional multiplex” example data, it looks like it takes a single layer adjacency matrix for the connectivity and a time-series for the functional data. What is it doing from there?
Dear Avalon,
-
So far the available multiplex pipelines take the time series (fMRI), connectivity matrices (DWI), or structural values (T1w, PET). In order to directly input your supra-adjacency matrices, you can use braph2 via the command line and create another file similar to MultiplexWU or MultiplexBU removing all the different functions applied to the input matrix B. You could also need to change the Importer for the pipeline that you want to use.
-
In this case it is because we re-use the importers from a single layer of connectivity and functional pipelines. Thus, from the functional data, the adjacency matrix is calculated by correlating the time series from the brain regions, whereas from the connectivity data we already take connectivity matrices, and just apply functions to make sure the adjacency matrix is symmetric, without self-connections and negative weights (can be chosen), and standardized.
Dear Avalon,
In our newest release (https://github.com/softmatterlab/BRAPH-2-Matlab/releases/tag/2.0.0.a3) we added connectivity multiplex pipelines, where the input is an adjacency matrix for each layer per subject.
To understand the structure of the input data, you can check the following tutorial http://braph.org/tutorials/general/connectivity-multiplex-data-format/ -
- You must be logged in to reply to this topic.