igaprobability: IgA Probability
In IgAScores: Score Taxon-Level IgA Binding in IgA-Seq Experiments

Description Usage Arguments Details Value Examples

View source: R/igaprobability.R

Calculate the IgA Postive/Negative Probability as described in Jackson et al. (2020, doi: 10.1101/2020.08.19.257501).

1	igaprobability(withinabund, gatesize, presortabund, nazeros = TRUE)

`withinabund`	Abundance of the bacteria in the IgA gate under investigation (can be calculated for either the pos/high or neg/low gating) (abundances should sum to 1 not as a %).
`gatesize`	The fraction of events in the flow cytometer within the gate under investigation (as a decimal fraction not a %).
`presortabund`	Abundance of the bacteria in whole sample before sorting by IgA (abundances should sum to 1 not as a %).
`nazeros`	Return NA if the within and tot abundances are both zero. Default is TRUE.

This function calculates the conditional probability that at bacteria will be sufficiently bound/not bound to immunoglobulin A (IgA) to end up in a given IgA gate based on its taxonomy. Calculated on one taxa for one sample.

This uses Bayes' theorem assuming:

That the relative abundance of a given taxon in the IgA gate under question represents the probability of being that taxa given that it is within the IgA gate (either high or low).
That the percentage of flow cytometery events binned into the IgA gate represents the probability of any bacteria being within the gate.
That the abundance of the given taxon in the input sample (or whole fraction) represent the probability that any bacteria is assigned to the taxon. If there is insufficient levels of a taxa in the whole fraction to account for its abundance in the IgA gate, the function assumes all of the taxa fall within this gate (i.e. a probability of 1).

Further details can be found in Jackson et al. (2020, doi: 10.1101/2020.08.19.257501).

A numeric value for the IgA Positive/Negative Probability (depending on the data used for 'withinabund' and 'gatesize') as defined in Jackson et al. (2020, doi: 10.1101/2020.08.19.257501).