neuroanatomyandconnectivity/gradient_analysis

gradient_analysis

Link to view this repository.

Requirements:

Python2.7:

standard packages

numpy
scipy
pandas
h5py
matplotlib
seaborn
gdist

Neuroimaging

nibabel
nilearn
neurosynth

Specific

surfdist
Already in ./gradient_data/src/: mapalign
SurfacePlotting: optional for visualization

Other software requirements

(Only for generating volumes from surfaces)
Connectome workbench: wb_command
Caret
FreeSurfer
FSL

gradient_analysis's Issues

memory error

Having issue reading in the HCP groupPCA data; get a memory error

f = 'HCP_S900_820_rfMRI_MSMAll_groupPCA_d4500ROW_zcorr.dconn.nii'    
dcon = np.tanh(nib.load(f).get_data()[:])

Do you have any suggestions on this?

(Looks like maybe this is what the hcp_corr repo is for; but can't quite tell if that gives same output as the above line)

Thx.

Error:jointplot() got multiple values for argument 'data'

Thank you professor Margulie for sharing this inspiring code！！
And I was trying to use this method on my own cor_matrix,size(148*148).
But I encountered a program error when creating the Scatterplot of first two dimensions(03_visualize_embeddings) :
g = (sns.jointplot('e1', 'e0',
data=df,
size=10, label='big', linewidth=0, marker='.', c=np.array(c), alpha=0.8,
ylim=[np.min(df['e0']) - 0.5, np.max(df['e0']) + 0.5],
xlim=[np.min(df['e1']) - 0.5, np.max(df['e1']) + 0.5],
stat_func=None).set_axis_labels('Gradient 2', 'Gradient 1'))
TypeError: jointplot() got multiple values for argument 'data'
I try to replace the ylim and xlim into specific number but it's still error.
Could you please tell me where the problem is?
Thanks again！

Missing file - hcp.tmp.lh.dscalar.nii

Hi again.

Getting following error:
nibabel.py3k.FileNotFoundError: No such file: 'gradient_data/templates/hcp.tmp.lh.dscalar.nii'

where is this file supposed to come from?
Ta.

macaque connectivitydata

Hi.

Thanks for the code.

I'm trying to run the macaque connectivity notebook; but it looks like some data files are missing

col = atlases.get(mat_name) mat = pd.read_csv(('gradient_data/macaque/%s_mat.txt' % mat_name), delimiter='\t', header=-1) name = pd.read_csv(('gradient_data/macaque/%s_name.txt' % mat_name), delimiter='\n', header=-1)

Could you please point me to where I can get these from?

Ta!

overlap of cortical and subcortical voxels in volumetric analysis

Summing the volumes in the volumetric portions of the analysis (meta-analyses) result in doubling the values within 998 voxels (in 2mm space), largely located with the hippocampus. This is due to the hippocampus being represented in both cortical and subcortical spaces in cifti space. The fix will affect the volumes that include subcortical regions on NeuroSynth as well as may have a minor impact on the results of the meta-analyses (due to hipppocampal values being doubled). All other results only represented cortical space and will remain unaffected.

where are .surf files from?

Notebook '03_visualize_embeddings.ipynb' uses a file called
gradient_data/templates/S900.L.midthickness_MSMAll.32k_fs_LR.surf

Where does this file come from?

Closest I know of is from a folder called HCP_S900_GroupAvg_v1.zip that I obtained from a slightly random internet location, which contains file S900.L.midthickness_MSMAll.32k_fs_LR.surf.gii

Seems that mapping between freesurfer and HCP surface spaces is a little more involved than one might like.

Do you have the commands used for this somewhere here?

Ta.

embded.compute_diffusion_map 's vectors is stuck into a cycle of positive and negative value

when I run the "emb, res =embded.compute_diffusion_map(aff, alpha=0.5, return_result = True)", which the "aff" is a 800×800 symmetrical matrix, it return the res with the "vectors" and "emb", if some value of the two array is positive for the first time, and then run the function again it will be negative with the almost same absolute value, I am quit confused about this problem and when I run a lot of matrix it will bring inconsistency, and it will be quite tricky to my work which is about the compute of gradient of human brain functional connectivity. I wish some help can be available for me.

Embedding maps availible?

Thanks for making availible the code for your very interesting paper!

I was hoping to get the embedding maps so that I can visualize them myself. I gather that I could create them using the notebooks here, but I would need the HCP connectome data and some processing time. Have you made the embedding maps that you show in the paper availible somewhere?

That is, I am looking for the files that will end up in gradient_data/embed and are shown here: https://github.com/NeuroanatomyAndConnectivity/gradient_analysis/blob/master/03_visualize_embeddings.ipynb. I think I specifically want the gradient_data/embedded/embedding_dense_emb.npy file, which I gather I could display on the 32k_fs_LR mesh?