search

wu-yc / scMetabolism

Quantifying metabolism activity at the single-cell resolution
BSD 3-Clause "New" or "Revised" License
115 stars 46 forks source link

how to handle large dataset(>65000 cells) in scMetabolism? #7

Open ybyOxidant opened 2 years ago

ybyOxidant commented 2 years ago

Hello! I am currently working with a dataset containing more than 60000 cells. Runing sc.metabolism or sc.metabolism.Seurat is blocked by the ERROR below. Error in asMethod(object) : Cholmod error 'problem too large' at file ../Core/cholmod_dense.c, line 102

I believe it will happen when the column number exceeds the limitation of as.matrix(). So now I break the dataset into 2 parts to process by scMeatbolism, then merging them for downstream analysis. But it's not elegant, and the internal parameter of scMetabolism seems not equal. Do you have a better solution?

wangshisheng commented 2 years ago

Hello!

I meet same problem, could you provide some help? Many thanks~~

abollol commented 2 years ago

Hello! me too. any solution?

honghh2018 commented 2 years ago

@wangshisheng @abollol same issue. how to solve this problem?

honghh2018 commented 2 years ago

What happen for this R package ? no one to maintain ?

wu-yc commented 2 years ago

This is limited by as.matrix(). You can split your datasets according to cell types (i.e. T cell, B cell, etc.) and perform the analysis one by one.