[DOC] Add data/ README.md

data/README.md

@@ -0,0 +1,16 @@
+# data
+
+This directory contains some summary data from the main analyses run in our manuscript.
+
+- [`./raw/gene_modules/oldham2008.xls`](./raw/gene_modules/oldham2008.xls): An Excel spreadsheet from the Supplementary Materials of [Oldham et al., 2008, *Nat Neuro*](https://doi.org/10.1038/nn.2207) denoting functional gene commmunity assignments.
+- [`./raw/dk/fslr32k/atlas-desikankilliany_space-fsLR_den-32k_hemi-{L,R}_dparc.label.gii`](./raw/dk/fslr32k): The surface Desikan-Killiany atlas from `abagen` projected to fsLR32k space for use in parcellating T1w/T2w data from the HCP.
+- [`./derivatives/pipelines.csv.gz`](./derivatives/pipelines.csv.gz): This file contains the analysis outputs from all ~750k pipelines assessed in the current work.
+  Each row represents a unique pipeline and columns contain information on the pipeline parameters and derived statistical estimates.
+- [`./derivatives/impact.csv`](./derivatives/impact.csv): This CSV contains the impact score for each parameter calculated separately for each analysis.
+- [`./derivatives/literature`](./derivatives/literature): Outputs of the pipelines reproduced from the literature.
+  HDF5 files contain the expression values as generated by `abagen`, and the `analysis.csv` file contains the statistical estimates from the different reproduced pipelines for the three analyses we examined.
+
+All data provided in the `derivatives` folder are provided for ease of reproducibility; note, however, that all of these results can be generated programmatically.
+(Refer to our [walkthrough](https://netneurolab.github.io/markello_transcriptome) for instructions on how to do this.)
+
+Note that if files provided through git LFS become unavailable (due to e.g., bandwidth limits) you can find duplicate copies on [OSF](https://osf.io/cdet7/).