preprocessing.Rmd

<br class="long"/>


# Preprocessing involves:

* Normalization
* Debarcoding
* Randomize 0 values between -1 and 0
* Transformations

<br>

## Normalization and Debarcoding

For normalization, this pipeline uses the <a href="https://www.bioconductor.org/packages/devel/bioc/html/CATALYST.html">R CATALYST package</a> which implements bead-based normalization as described by Finck et al. 2013 ^1^. 


Here, identification of bead-singlets (used for normalization), as well as of bead-bead and cell-bead doublets (to be removed) is automated as follows:

1. beads are identified as events with their top signals in the bead channels
2. cell-bead doublets are remove by applying a separation cutoff to the distance between the lowest bead and highest non-bead signal
3. events passing all vertical gates defined by the lower bounds of bead signals are removed (these include bead-bead and bead-cell doublets)
4. bead-bead doublets are removed by applying a default median±5 rule to events identified in step 2. The remaining bead events are used for normalization.

<br>

For debarcoding, an implementation of the single-cell deconvolution algorithm described by Zunder et al.2015 ^2^ was used from the CATALYST package.

<br>


### Randomize 0 values between -1 and 0 { #rand }
Before performing any transformation 0 values are randomized between -1 and 0. This helps in visualizing the marker intensities as it removes the peak at 0.

<br>

## Transformations

There are 4 options for transformations. These are implemented by functions in R packages  <a href="https://www.bioconductor.org/packages/devel/bioc/html/flowCore.html">flowCore</a>  and <a href="https://www.bioconductor.org/packages/devel/bioc/html/cytofkit.html">Cytofkit</a>.

Transformation| Description
---------|------------------------------------------
logicl|Logicle transformation creates a subset of biexponentialTransform hyperbolic sine transformation functions.
arcsinh| hyperbolic arcsine (arcsinh) transformation
cytofAsinh|Inverse hyperbolic sine transformation (arsinh) with a cofactor of 5, reduce noise from negative values
none| No transformation is performed


Each of these transformation takes in different input parameters for which default values are provided but in some cases, flexibility is needed to play around with these parameters to get a more appropriate transformation.


### References

1. R. Finck et al., Normalization of mass cytometry data with bead standards. Cytometry Part A. 83A, 483–494 (2013).
2. E. R. Zunder et al., Palladium-based mass tag cell barcoding with a doublet-filtering scheme and single-cell deconvolution algorithm. Nat. Protocols. 10, 316–333 (2015).