Error in ge %*% M.AV : Cholmod error 'problem too large' at file ../Core/cholmod_dense.c, line 102 Calls: makeSeuratMC -> supercell_GE -> %*% -> %*% In addition: Warning messages: 1: In asMethod(object) : sparse->dense coercion: allocating vector of size 19.0 GiB 2: In asMethod(object) : sparse->dense coercion: allocating vector of size 2.5 GiB Execution halted

[christophechu]

Error in ge %% M.AV :
Cholmod error 'problem too large' at file ../Core/cholmod_dense.c, line 102
Calls: makeSeuratMC -> supercell_GE -> %% -> %*%
In addition: Warning messages:
1: In asMethod(object) :
sparse->dense coercion: allocating vector of size 19.0 GiB
2: In asMethod(object) :
sparse->dense coercion: allocating vector of size 2.5 GiB
Execution halted

Too Large for supercell_GE

[leonardHerault]

Hello,

You can try to split your ge single cell matrix in several groups and the metacell membership vector accordingly and then bind the resulting metacell ge matrices.

If you are working with Seurat object you can also try to use the AggregateExpression Seurat function by adding in the seurat object a metadata column containing the metacell membership vector.

Best

LH