Mouse / human control of cross-contaminations in single-cell ATAC-seq on Fluidigm C1
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
161026_M00528_0239_000000000-ANY8K_files/figure-html
.gitignore
161026_M00528_0239_000000000-ANY8K.Rmd
161026_M00528_0239_000000000-ANY8K.dedupStats.log
161026_M00528_0239_000000000-ANY8K.dedupStatsQ20.log
161026_M00528_0239_000000000-ANY8K.html
161026_M00528_0239_000000000-ANY8K.md
161026_M00528_0239_000000000-ANY8K.redundancy.log
161026_M00528_0239_000000000-ANY8K.redundancy2.log
161026_M00528_0239_000000000-ANY8K.zMeta.tsv
161026_M00528_0239_000000000-ANY8K.zUpload.tsv
1772-123-148_cell-count.xlsx
LICENSE
README.md
scATAC-mouse-human-control.Rproj
sequenceAlignment.md

README.md

C1 ATAC-seq cross-contamination control

Mouse / human control of cross-contaminations in single-cell ATAC-seq on Fluidigm C1.

The purpose of this repository is to announce, document, and track issues and contributions related to a single-cell ATAC-seq dataset that we created to measure and control cross-contaminations. Please feedback with the issue tracker here on GitHub, or with the Biostars forum.

To rule out or quantify the amount of cross-contamination in single-cell ATAC-seq libraries prepared on the C1 platform with the protocol available on ScriptHub, we performed the control experiemnts described below, loading a mixture of mouse and human cells.

We used medium-sized IFCs that had the original design prone to doublet capture. The chambers were imaged at multiple focal planes and the images are available in our single-cell data integration platform. The sequencing data is available on Zenodo, at the links displayed below for each experiment.

Experiment 1: mouse / human cell mixture

In the first experiment, we loaded mouse and human cells together, after staining them with different calceins.

  • Hep G2 (human), stained witn green calcein. 12.8 μm diameter in average.
  • Hepa 1-6 (mouse), stained wiht red calcein. 13.3 μm diameter in average.
  • IFC ID 1772-123-148 (old design)
  • MiSeq run ID: 161026_M00528_0239_000000000-ANY8K https://doi.org/10.5281/zenodo.263694

This experiment gives an estimate of the cross-contaminations in situations where the cells are robust and the possibility of damage before loading was limited.

Experiment 2: FACS-sorted mouse and human cells with / without washing

In the second experiment, we FACS-sorted the cells before loading, thus adding one potential source of stress and damage. Nevertheless, we used gentle FACS conditions (nozzle size of 100 μm), to reflect the actual settings that would be recommended when working with fragile cells. We loaded the cells with or without washing them after FACS-sorting. 48 samples (single cells, empty chambers, doublets) from each C1 run were multiplexed in one MiSeq run.

Cell preparation:

  • 2~5,000,000 cells collected in PBS and then FACS-sorted.

Brief experimental summary:

  • Same cell types and staining as in experiment 1.
  • sorting (collection into native medium)
  • cell count -> load in first C1 machine (IFC ID 1772-123-155, old design)
  • 300 × g, 5 min, then discard supernatant
  • resuspend with Cell Wash buffer (for C1)
  • cell count -> load in second C1 machine (IFC ID 1772-123-158 old design)
  • 250 cells/ul preparation (10,000 - 30,000 cells needed)
  • MiSeq run ID: 170116_M00528_0252_000000000-B3BKB https://doi.org/10.5281/zenodo.263695

Microscope pictures of the capture chambers

Data analysis

Our data analysis is ongoing. Here is a description of our alignment strategy. Comments and suggestions are welcome (ideally via Biostars).