02-quality-control-G-saturation.Rmd

```{r parameters-and-defaults, include = FALSE}
module <- "scRNAseq"
section <- "quality_control"
```

```{r parameter-merge, include = FALSE}
local_params <- module %>%
  options() %>%
  magrittr::extract2(module) %>%
  magrittr::extract2(section) %>%
  ReporteR.base::validate_params(parameters_and_defaults)
```

```{r scRNAseq-quality-control-G-saturation-checks}
assertive.sets::assert_is_subset(local_params$features, colnames(SummarizedExperiment::colData(object)))
```

### Sequencing saturation

Sequencing is called *saturated* when generating more sequencing output from a cDNA library does not substantially increase the number of detected features in a sample. Since the number of detected features can act as a technical confounder, and thereby drive substructure in the data, it is advisable to aim for a saturated sequencing by either adding more sequencing output or decreasing the number of samples until saturation is achieved (Figure \@ref(fig:scRNAseq-quality-control-G-saturation-figure)). [@zhang_one_2018] gives advise on how to choose the optimal cell number given a fixed sequencing budget. 

```{r scRNAseq-quality-control-G-saturation-figure, warning = FALSE, message = FALSE, echo = FALSE, fig.cap = "Scatterplot between sequencing depth and detected features. The dashed red line indicates a smooth trend that has been fit to the data using a general additive model. Saturation occurs when the trend line becomes flat."}
object_filtered %>%
  scater::plotColData(y = "total_features", x = paste0("total_", local_params$assay), colour_by = local_params$features[1], show_se = FALSE, show_smooth = TRUE) +
  ggplot2::xlab("Sequencing depth") +
  ggplot2::ylab("Detected features")
```