Multiple occurrences gene count quantified in single cell gene expression. Which unique gene count to consider for differential expression among samples

Question

[singlecell]

Hi.

The single cell gene expression for the cells had been perfomed using icell8 platform and the data is in this format

Type    gene_id gene_name   sample  bx  READNAME    um                       
exon    ENSG00000000457 SCYL3   Miao    ACGAAGAGAGG 183 1  
exon    ENSG00000000457 SCYL3   Miao    CTGGCTGGTTG 108 3  
exon    ENSG00000000457 SCYL3   Miao    GAACCTCTATG 1   1  
exon    ENSG00000000457 SCYL3   Miao    GTCTATGGACC 321 4  
exon    ENSG00000000457 SCYL3   Miao    GTCTGCGCTAT 192 2  
exon    ENSG00000000457 SCYL3   Miao    GTTGCCGTCAG 51  2   
exon    ENSG00000000457 SCYL3   Miao    TCGGTATACCA 500 7  
exon    ENSG00000000460 C1orf112    Miao    AACCGCGCCAT 10  1  
exon    ENSG00000000460 C1orf112    Miao    AACCGGCAAGC 8   2.  
exon    ENSG00000000460 C1orf112    Miao    AACCTTACGGC 310 4   
exon    ENSG00000000460 C1orf112    Miao    AACGATCAGTT 297 2

The multiple genes like SCYL3 and the barcode ACGAAGAGAGG which corresponds to well e.g searching it in well list means
SYCL3 with barcode ACGAAGAGAGG belongs to sample S1. samples S1,S2 and S4 is annotated from well list e.g

("1462" "ACGAAGAGAGG" 58 12 "S1" "C2" NA NA)

Do we need data for DE analysis in the format listed below?

gene_name gene_id S1 S2 S4
SYCL3 ENSG00000000457 183  1 108

But when we are searching for GTCTATGGACC barcode and it belongs to sample S4, so now there are 2 counts for SYCL3 in sample S4 and this is similar for other genes where multiple gene entries are found in samples.

Sample well list file below: We already know Column and Rows belong the which sample e.g S1, S2 or S4

Row Col Candidate   For dispense    Sample  Barcode State   Cells1  Cells2  Signal1 Signal2 Size1   Size2   Integ Signal1   Integ Signal2   Circularity1    Circularity2    Confidence  Confidence1 Confidence2 Dispense tip    Drop index  Global drop index   Source well Sequencing count    Image1  Image2
41  56  True    True    Miao    AACGAGTATAT Good    1   0   7113        45      320085      1       0.9506  0.97    0.98    1   158 201 A1      Pos81_1-Hoechst_G10.tif Pos81_4-Rhodamine_G10.tif
33  30  True    True    Miao    CTCCGACCTAG Good    1   0   6654        49      326046      1       0.9220821   0.9409  0.98    7   59  71  D1      Pos65_1-Hoechst_F06.tif Pos65_4-Rhodamine_F06.tif

If I'm correct, for differential expression analysis of genes among samples , we need to have row having unique genes and counts corresponding to that genes in the sample. Which count to consider? Please correct me if I'm not correct!!

Appreciate your help!!

Best

Ankush

[Ram]

Please use the formatting bar (especially the code option) to present your post better. I've done it for you this time.

[singlecell]

Thanks for your help !!!!