quadbio / simspec

Calculation of Reference/Cluster Similarity Spectrum (RSS/CSS)
30 stars 3 forks source link

is it possible to launch a CSS from a MNN? #8

Closed tsterv closed 1 year ago

tsterv commented 1 year ago

Hello,

I would like to use CSS to aggregate my samples on a single matrix without going through the cellranger aggregation (batch-effect problem) but I can't do it. I had previously used FastMNN to do it, but a PhD student advised me to use simspec for my data (hiPSC-CMs in different differenciation states).

Here is the code I used (each DCXXXX is a Seurat Object) :

Data <- cluster_sim_spectrum( object = list(DC2102, DC2149, DC2093, DC2158,
                                                 DC3626, DC3720_j8,DC3720_j30, DC3725,DC3785,DC3664_j8,DC3664_j30,DC3724_j8,DC3724_j30,DC3788_j8,DC3788_j30,DC3625,DC3662,DC3783,DC3784,DC3795,
                                                 DC3388, DC3389, DC3390, DC3391, DC3392, DC3393, DC3411, DC3412, DC3413, DC3414, DC3415, DC3416), 
                              label_tag = "orig.ident",
                              cluster_resolution = 0.4,
                              labels = "orig.ident",
                              corr_method = "pearson",
                              spectrum_type = "corr_kernel")

Can i used the seurat object from the fastMNN agregate object or i need to use my samples ?

Any solution would be welcome !

Thank you