A: Data resources and values for single cell genomics

Vincent J. Carey, stvjc at channing.harvard.edu

March 27, 2023

Source: vignettes/A_Data.Rmd

The SingleCellExperiment class

sce

sce

Bioconductor’s scRNAseq resource

The scRNAseq package includes functions that will retrieve exemplary datasets, place them in a user-defined cache (so that retrievals need not be repeated), and deliver a representation into the workspace.

Enumerating the functions and their sources in pubmed 

data(curated_scrnaseq_table)
datatable(curated_scrnaseq_table, escape=FALSE)

Searchable abstracts (specially curated) 

data(curated_abstdf)
curated_abstdf$pmids = build_anchors(curated_abstdf$pmids)
datatable(curated_abstdf, escape=FALSE)

Using it 

library(scRNAseq)
zld = ZilionisLungData()
zld
## class: SingleCellExperiment 
## dim: 41861 173954 
## metadata(0):
## assays(1): counts
## rownames(41861): 5S_rRNA 5_8S_rRNA ... snosnR66 uc_338
## rowData names(0):
## colnames(173954): bcIIOD bcHTNA ... bcELDH bcFGGM
## colData names(16): Library Barcode ... used_in_NSCLC_non_immune
##   used_in_blood
## reducedDimNames(5): SPRING_NSCLC_all_cells
##   SPRING_NSCLC_and_blood_immune SPRING_NSCLC_immune
##   SPRING_NSCLC_non_immune SPRING_blood
## mainExpName: NULL
## altExpNames(0):
plot(reducedDim(zld), pch=".", col=factor(zld$`Major cell type`))

The projection is based on Klein’s SPRING.

Exercises: Supply a legend. Use more modern visualization. Investigate second-level cell type assignments.

The Human Cell Atlas 

metadata  = get_metadata()
metadata
## # Source:   table</home/stvjc/.cache/R/CuratedAtlasQueryR/metadata.parquet> [?? x 56]
## # Database: DuckDB 0.7.1 [stvjc@Linux 5.15.0-67-generic:R 4.3.0/:memory:]
##    .cell          sample_id_db .sample .sample_name assay assay_ontology_term_id
##    <chr>          <chr>        <chr>   <chr>        <chr> <chr>                 
##  1 AAACCTGAGAGAC… 8a0fe092868… 5f20d7… D17PrPzF_Vi… 10x … EFO:0009899           
##  2 AAACCTGAGTTGT… 8a0fe092868… 5f20d7… D17PrPzF_Vi… 10x … EFO:0009899           
##  3 AAACCTGCAGTCG… 02eb2ebcb5f… 5f20d7… D17PrPzF_Vi… 10x … EFO:0009899           
##  4 AAACCTGCAGTTC… 02eb2ebcb5f… 5f20d7… D17PrPzF_Vi… 10x … EFO:0009899           
##  5 AAACCTGGTCTAA… 8a0fe092868… 5f20d7… D17PrPzF_Vi… 10x … EFO:0009899           
##  6 AAACCTGTCGTAC… 8a0fe092868… 5f20d7… D17PrPzF_Vi… 10x … EFO:0009899           
##  7 AAACCTGTCTTGT… 8a0fe092868… 5f20d7… D17PrPzF_Vi… 10x … EFO:0009899           
##  8 AAACGGGAGTACG… 8a0fe092868… 5f20d7… D17PrPzF_Vi… 10x … EFO:0009899           
##  9 AAACGGGAGTAGG… 02eb2ebcb5f… 5f20d7… D17PrPzF_Vi… 10x … EFO:0009899           
## 10 AAACGGGAGTGGG… 8a0fe092868… 5f20d7… D17PrPzF_Vi… 10x … EFO:0009899           
## # ℹ more rows
## # ℹ 50 more variables: file_id_db <chr>, cell_type <chr>,
## #   cell_type_ontology_term_id <chr>, development_stage <chr>,
## #   development_stage_ontology_term_id <chr>, disease <chr>,
## #   disease_ontology_term_id <chr>, ethnicity <chr>,
## #   ethnicity_ontology_term_id <chr>, file_id <chr>, is_primary_data.x <chr>,
## #   organism <chr>, organism_ontology_term_id <chr>, …
metadata |> dplyr::count()
## # Source:   SQL [1 x 1]
## # Database: DuckDB 0.7.1 [stvjc@Linux 5.15.0-67-generic:R 4.3.0/:memory:]
##          n
##      <dbl>
## 1 28975366

Filtering cells by donor, assay, tissue, cell annotation:

single_cell_countsALC = 
    metadata |>
    dplyr::filter(
        ethnicity == "African" &
        stringr::str_like(assay, "%10x%") &
        tissue == "lung parenchyma" &
        stringr::str_like(cell_type, "%CD4%")
    ) |>
    get_SingleCellExperiment()
## ℹ Realising metadata.
## ℹ Synchronising files
## ℹ Reading files.
## ℹ Compiling Single Cell Experiment.
single_cell_countsALC
## class: SingleCellExperiment 
## dim: 36229 1571 
## metadata(0):
## assays(2): counts cpm
## rownames(36229): A1BG A1BG-AS1 ... ZZEF1 ZZZ3
## rowData names(0):
## colnames(1571): ACAGCCGGTCCGTTAA_F02526_1 GGGAATGAGCCCAGCT_F02526_1 ...
##   TACAACGTCAGCATTG_SC84_1 CATTCGCTCAATACCG_F02526_1
## colData names(56): sample_id_db .sample ... n_tissue_in_cell_type
##   original_cell_id
## reducedDimNames(0):
## mainExpName: NULL
## altExpNames(0):
head(colData(single_cell_countsALC)[,1:5])
## DataFrame with 6 rows and 5 columns
##                                     sample_id_db                .sample
##                                      <character>            <character>
## ACAGCCGGTCCGTTAA_F02526_1 33cdeb84ae1462d723c1.. 4fc10a6b85e5fa688b25..
## GGGAATGAGCCCAGCT_F02526_1 33cdeb84ae1462d723c1.. 4fc10a6b85e5fa688b25..
## TCTTCGGAGTAGCGGT_F02526_1 33cdeb84ae1462d723c1.. 4fc10a6b85e5fa688b25..
## CCTTACGAGAGCTGCA_F02526_1 33cdeb84ae1462d723c1.. 4fc10a6b85e5fa688b25..
## ATCTACTCAATGGAAT_F02526_1 33cdeb84ae1462d723c1.. 4fc10a6b85e5fa688b25..
## CATCAGACACCGTTGG_F00409_1 c7f14e86df84af9ad171.. 9ef5eab00316c489a363..
##                                     .sample_name       assay
##                                      <character> <character>
## ACAGCCGGTCCGTTAA_F02526_1 VUHD92___lung parenc..   10x 5' v1
## GGGAATGAGCCCAGCT_F02526_1 VUHD92___lung parenc..   10x 5' v1
## TCTTCGGAGTAGCGGT_F02526_1 VUHD92___lung parenc..   10x 5' v1
## CCTTACGAGAGCTGCA_F02526_1 VUHD92___lung parenc..   10x 5' v1
## ATCTACTCAATGGAAT_F02526_1 VUHD92___lung parenc..   10x 5' v1
## CATCAGACACCGTTGG_F00409_1 VUHD69___lung parenc..   10x 5' v1
##                           assay_ontology_term_id
##                                      <character>
## ACAGCCGGTCCGTTAA_F02526_1            EFO:0011025
## GGGAATGAGCCCAGCT_F02526_1            EFO:0011025
## TCTTCGGAGTAGCGGT_F02526_1            EFO:0011025
## CCTTACGAGAGCTGCA_F02526_1            EFO:0011025
## ATCTACTCAATGGAAT_F02526_1            EFO:0011025
## CATCAGACACCGTTGG_F00409_1            EFO:0011025

Reaching across: data available in scvi-tools or scanpy

Use scviHelper() or scanpyHelper() apps in scviR.