Some problem about GeneIntegrade Matrix

[yangxueqianR]

First , thank you very much for developing such a useful analysis tool. There is some confusion about the GeneIntegrade Matrix. I get matrix used getMatrixFromProject function.

mat = getMatrixFromProject(proj,useMatrix="GeneIntegrationMatrix")
ArchR logging to : ArchRLogs/ArchR-getMatrixFromProject-735e5764dfe4-Date-2023-03-03_Time-11-13-02.log
If there is an issue, please report to github with logFile!
2023-03-03 11:14:52 : Organizing colData, 1.83 mins elapsed.
2023-03-03 11:14:54 : Organizing rowData, 1.864 mins elapsed.
2023-03-03 11:14:54 : Organizing rowRanges, 1.864 mins elapsed.
2023-03-03 11:14:54 : Organizing Assays (1 of 1), 1.864 mins elapsed.
2023-03-03 11:17:45 : Constructing SummarizedExperiment, 4.723 mins elapsed.
2023-03-03 11:17:48 : Finished Matrix Creation, 4.775 mins elapsed.

And then I looked at the dimensions of the matrix.

dim(mat)
[1] 11697 50637

His cell number is consistent with that of ATAC, less than that of RNA. The number of genes is greater than the number of gene overlap between GeneScore and RNA.
Therefore, I took out the matrix in the Seurat object and corresponded the predictedCells with the corresponding genes, and found that the two were inconsistent

sc = readRDS("all_article_celltype.rds")
sc = sc[rowData(mat)$name,proj$predictedCell]
expr1 = assay(mat)
expr2 = sc@assays$RNA@data

expr1[1:4,1:4]
4 x 4 sparse Matrix of class "dgCMatrix"
     24hpf_1#24hpf_1_BC00224_N06 24hpf_1#24hpf_1_BC00014_N03
[1,]                  83.2416934                 80.87149456
[2,]                   0.0811268                  .         
[3,]                   0.3215362                  0.25338820
[4,]                   0.2523174                  0.05113611
     24hpf_1#24hpf_1_BC00075_N02 24hpf_1#24hpf_1_BC00171_N03
[1,]                 79.76752139                 98.10917315
[2,]                  0.06844169                  0.05498361
[3,]                  0.39032410                  0.17883260
[4,]                  0.38505563                  0.39857684

expr2[1:4,1:4]
4 x 4 sparse Matrix of class "dgCMatrix"
        6h_3 CELL4645_N1 _ 3h1_CELL1337_N1_3h1 3h1_CELL1337_N1_3h1
rpl24             1.348236           2.8083793           2.8083793
cep97             .                  1.1472063           1.1472063
eed               .                  1.3047392           1.3047392
hikeshi           .                  0.9601302           0.9601302
        3h1_CELL1197_N1_3h1
rpl24             3.0070604
cep97             0.6991253
eed               1.1065752
hikeshi           .        

So I want to know how to understand the GeneIntegrade Matrix.

[rcorces]

Probably because there is normalization applied.

ArchR/R/RNAIntegration.R

Line 280 in 381842c

seuratRNA <- NormalizeData(object = seuratRNA, verbose = FALSE)

If you want additional help, please create a reproducible example from the tutorial data rather than showing your own data