sciplex3 count matrix gene indices

[yyou1996]

Thank you for the great resource.

I notice that for sciplex3 data (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSM4150378)

• The count matrix (GSM4150378_sciPlex3_A549_MCF7_K562_screen_UMI.count.matrix.gz) contains gene indices ranged in [1, 110981];
• The gene annotations (GSM4150378_sciPlex3_A549_MCF7_K562_screen_gene.annotations.txt.gz) contains 110983 genes.

Do I understand correctly that the indices correspond to the row number of genes, and the last two genes are out of consideration?

[Fazel-AVB]

Hi @yyou1996,

I have the same issue with sciplex3. Could you please let me know if you found the answer to your question?

Best,
Fazel

[yyou1996]

Not really from here. I end up to use a version of processed data from someone else and still have no clue how it works here. You can search for keywords of causal inference, single cells, sciplex and you can find some repo providing the processed data.

I will still appreciate a lot if I can know how it works on the raw data ! :)

[Fazel-AVB]

Hi @yyou1996,

Thanks for your response. Do you mean this paper? Causal identification of single-cell experimental perturbation effects with CINEMA-OT

Because when I checked it, I realised they have used the sciplex4, and in the sciplex4, the number of genes in the metadata file is equal to the highest gene index in the count matrix (58347).