how to calculate gene scores for healty tissues ？

[bitcometz]

helllo， there are no gene scores for healty tissue, which is shown in your supplementary tables:

celltypes | traits | Escore | pEscore -- | -- | -- | -- Bcells (1.0%) (Healthy) | Inflammatory bowel disease | 5.919 | 1.82E-05 Bcells (0.4%) (Disease Progression) | Inflammatory bowel disease | 7.063 | 3.68E-05 Dendritic cells (1.1%) (Healthy) | Inflammatory bowel disease | 8.524 | 2.86E-07 Dendritic cells (0.3%) (Disease Progression) | Inflammatory bowel disease | 6.673 | 0.000162

I’m wondering whether the calculation method of healthy tissue is the same as that of diseased tissue, but the comparison is reversed by using the scanpy:

sc.tl.rank_genes_groups

Best

[kkdey]

The healthy gene program is from the xavier_colon tissue set of gene scores on the website. The calculation of the healthy and diseae progression are a little different, you can check the Methods section of our paper.

[bitcometz]

Thanks for your reply !!!

But there not 'xavier_colon' : (datapath + '/colon/healthydata.h5ad', 'subject') in the following website:

https://alkesgroup.broadinstitute.org/LDSCORE/Jagadeesh_Dey_sclinker/scdata/

Is it the subset from the ucdata.h5ad with the conditions as 'healthy'?

Best

[kkdey]

Yes xavier_colon healthy cell tyoe program is derivd from ucdata.h5ad file. So, you ned to restrict yourself to healthy samples in that h5ad file and then compute the DE

[kkdey]

@bitcometz All healthy cell type programs are in https://alkesgroup.broadinstitute.org/LDSCORE/Jagadeesh_Dey_sclinker/gene_scores/cell_type_programs/ Alzheimers: alzheimers_brain UC: xavier_colon MS: msbrain Lung fibrosis: krasnow_lung Lung asthma: asthma_healthy Lung covid: idoamit_healthy

[bitcometz]

Thanks ！！！