Tutorial for preparing and checking input data.¶

collectNET supports four file input formats for online inference: txt file (.txt or .txt.gz); R data formatted file (.rds); Anndata data file (.h5ad); Seurat object data formatted file (.h5Seurat).¶

Please perform the following checks before inputting each file type.¶

1. txt file (.txt or .txt.gz).¶

Here collectNET needs both single-cell raw count matrix and cell annotation data for txt-format input.¶

Image
In [1]:
library(Seurat)
genecount <- as.matrix(read.table('txt_example.txt', header = TRUE, sep = ',')) ## ensure that the content in the txt file is comma-separated
genemeta <- read.csv('cell_anno_example.csv')
ser.obj <- CreateSeuratObject(genecount, meta.data=genemeta)
print(ser.obj)

Attaching SeuratObject

Seurat v4 was just loaded with SeuratObject v5; disabling v5 assays and
validation routines, and ensuring assays work in strict v3/v4
compatibility mode


An object of class Seurat 
5000 features across 1308 samples within 1 assay 
Active assay: RNA (5000 features, 0 variable features)
 2 layers present: counts, data

2. R data formatted file (.rds)¶

In [2]:
library(Seurat)
ser.obj <- readRDS('seurat_example.rds')
print(ser.obj)

An object of class Seurat 
21005 features across 1308 samples within 1 assay 
Active assay: RNA (21005 features, 0 variable features)
 2 layers present: counts, data

3. Anndata data file (.h5ad)¶

In [3]:
library(MuDataSeurat)
ser.obj <- ReadH5AD('h5ad_example.h5ad')
print(ser.obj)

An object of class Seurat 
22315 features across 6022 samples within 1 assay 
Active assay: RNA (22315 features, 0 variable features)
 2 layers present: counts, data

4. Seurat object data formatted file (.h5Seurat)¶

In [4]:
library(SeuratDisk)
ser.obj <- LoadH5Seurat('h5Seurat_example.h5Seurat')
print(ser.obj)

Registered S3 method overwritten by 'SeuratDisk':
  method            from  
  as.sparse.H5Group Seurat

Validating h5Seurat file

Initializing RNA with data

Adding counts for RNA

Adding miscellaneous information for RNA

Adding command information

Adding cell-level metadata

Adding miscellaneous information

Adding tool-specific results


An object of class Seurat 
5000 features across 6022 samples within 1 assay 
Active assay: RNA (5000 features, 0 variable features)
 2 layers present: counts, data

Check if a proper single-cell raw count data is available¶

In [5]:
print(ser.obj)

An object of class Seurat 
5000 features across 6022 samples within 1 assay 
Active assay: RNA (5000 features, 0 variable features)
 2 layers present: counts, data
In [6]:
print(GetAssayData(ser.obj, assay="RNA")[1:10, 1:10])
### select the first 10 rows and fist 10 columns as an example output

10 x 10 sparse Matrix of class "dgCMatrix"

  [[ suppressing 10 column names ‘AdultLung_2.CTCGCAAACCTAGGCTGC’, ‘AdultLung_2.CTGTGTCTCCATACGTTG’, ‘AdultLung_2.CTCGCAACGTTGGTAATG’ ... ]]


                                       
SFTPC     .  .  10 38  .  . 92  6 20  1
SFTPB     .  .   2  7  .  . 24  1  5  4
FTL       . 69 124 21 93  3  3 23  4 13
MT-RNR2 106 27  29 96 18 81 37 54 67  .
SFTPA2    2  3   1  2  1  . 10  .  1  1
TMSB4X   30 61 103 21 81 37 19 19 31  .
SCGB1A1   .  .   .  .  .  .  .  .  .  .
SFTPA1    1  .   .  4  .  . 14  2  1  .
B2M      25 24  46 15 54 39 20 18 60 14
TPT1     15  8   8 11  4  9 23  7  9  1

Check if cell names in the gene expression matrix and the metadata are the same¶

In [7]:
print(all(colnames(ser.obj@assays$RNA) == ser.obj@meta.data$Cell_id))

[1] TRUE

Check if column 'Celltype' exists in metadata¶

In [8]:
print(any(grepl('Celltype', colnames(ser.obj@meta.data))))

[1] TRUE