Cell type

[paolo maccallini]

Hi,

I have a problem with cell type analysis from the results of SNP2GENE job 685538. In particular, the CellType jobs are  685557 and  685645.

Thanks

[Tanya Phung]

For job  685557, there was a problem with the dataset 550_Han_Human_2020_AdultCerebellum_level2.txt. Specifically, these 2 cell type labels exist in the original datasets: Fetal_neuron and Fetal_Neuron, most likely due to typo from the original's dataset. This caused an issue with MAGMA. For the time being, I re-labelled Fetal_Neuron to  Fetal_Neuron_2 to bypass the error, but I can't tell you if these 2 labels are actual the same and whether it should be merged. Therefore, caution should be taken into consideration when using this dataset. 
Additionally, I noticed that you selected multiple levels for the same datasets. This is not how we intend FUMA cell type to be run. Please read here: 

For job  685645, MAGMA could not be performed with this dataset: DropViz_all_level1. I suggest you remove this dataset. If you want to figure out why, you could download the processed scRNAseq data (available in the Download tab), then run MAGMA on your environment to see why it failed with this dataset. 

Best,
Tanya