# multiGIviewer
An interactive R {shiny} web application built with the {golem}
framework to view multi-screen genetic interaction data
# Contents
- [Usage](#usage)
- [Inputs](#inputs)
- [Outputs](#outputs)
# Usage
Watch this short demo on how to use `multiGIviewer` to view replicated
genetic interactions (qGI scores) across two or more CRISPR screens of
interest.

# Inputs
| Parameters | Values | Properties |
|-------------------------|--------------------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| **Dataset** | GIN\_20210406, GIN\_20210107, GIN\_20201129, CHEM\_20210218, CHEM\_20210218\_GIN\_20210406 | qGI scores for 280 (GIN\_20210406), 257 (GIN\_20210107), 247 (GIN\_20201129), 68 (CHEM\_20210218), 348 (CHEM\_20210218) & (GIN\_20210406) screens \[*required*; default GIN\_20210406\] |
| **Screens** | screens IDs | list of all screens in selected dataset; two or more screens need to be selected to view replicated interactions \[*required*\] |
| **Media conditions(s)** | min, rich, pyro and/or DMSO | media type used to define wildtype single mutant fitness; more than one can be selected if screens were done in different media types \[*required*\] |
| **FDR threshold** | value from 0 to 1 | threshold to define significant genetic interactions; interactions must pass threshold in at least two screens to be shown \[*required*; default 0.2\] |
| **Positive colour** | HEX colour codes | colour to fill darkest positive points \[*optional*; default `#FAE057`\] |
| **Negative colour** | HEX colour codes | colour to fill darkest negative points \[*optional*; default `#61C2FA`\] |
| **Plot labels** | gene symbols | comma separated (character sensitive) list of genes to label on plot \[*optional*; default top ten positive and negative interactions\] |
| **Label type** | Text or Padded box | method to draw plot labels; padded box wraps text in a white box to better visualize labels \[*optional*; default Text\] |
| **Reference line(s)** | y=x, x=0 and/or y=0 | selection of reference lines to draw on plot \[*optional*\] |
# Outputs
#### Plot
| Elements | Properties |
|---------------------------------------------|------------------------------------|
| **Fitness HAP1 wildtype \[LFC\] (x-axis)** | corresponds to `mean_wtLFC` column |
| **Fitness HAP1 knockout \[LFC\] (y-axis)** | corresponds to `mean_koLFC` column |
| **Genetic interaction in n screens (fill)** | corresponds to `n_sig` column |
| **Mean qGI score (size)** | corresponds to `mean_qGI` column |
#### Table
| Columns | Properties |
|-----------------|-----------------------------------------------------------------------------------------|
| **gene** | gene interaction |
| **mean\_qGI** | mean qGI score of interaction across selected screens |
| **min\_FDR** | minimum qGI FDR value of interaction across selected screens |
| **mean\_wtLFC** | mean wildtype dropout effect (log2-foldchange) across selected media condition(s) |
| **mean\_koLFC** | mean knockout dropout effect (log2-foldchange) across selected screens |
| **n\_sig** | number of times positive or negative interaction is significant across selected screens |
| **screen\_sig** | list of screens where interaction is significant |