dashboard.Rmd

---
title: "Immunogenicity - Tiered Approuch to Assess ADA	Positive Samples"
output: 
  flexdashboard::flex_dashboard:
    orientation: rows
    vertical_layout: fill
params:
  screening: "Sample_ADA_Data_05062017_Screening.csv"
  confirmatory: "Sample_ADA_Data_05062017_Confirmatory.csv"
  screening_cut_point: 200
  confirmatory_cut_point: 20
---


```{r setup, include=FALSE}
library(flexdashboard)
library(tidyverse)
library(metricsgraphics)
library(RColorBrewer)

screening <- read_csv(params$screening)
confirmatory <- read_csv(params$confirmatory)

samples <- screening %>%
  left_join(confirmatory, by = "Sample_Number") %>%
  mutate(Signal_Response_Difference = Signal_Response_No_Drug - Signal_Response_Drug)  %>% 
  mutate(Signal_Response_Divide = Signal_Response_Difference / Signal_Response_No_Drug)  %>%
  mutate(Percent_Signal_Inhibition_Drug = Signal_Response_Divide * 100) %>%
  mutate(Screening_Result_Drug = ifelse(Signal_Response_No_Drug > params$screening_cut_point, "Positive", "Negative")) %>%
  mutate(Confirmatory_Result_Drug = ifelse(Percent_Signal_Inhibition_Drug > params$confirmatory_cut_point, "Positive", "Negative"))   %>%
  mutate(True_Positive = Screening_Result_Drug == Confirmatory_Result_Drug) %>%
  mutate_if(is.integer, as.numeric) 


samples
```

Row
-----------------------------------------------------------------------

### Observations
```{r}
valueBox(nrow(samples), "Observations", icon = "fa-flask")
```

### Screening
```{r}
valueBox(params$screening_cut_point, "Screening Cut Point", icon = "fa-filter")
```

### Confirmatory
```{r}
valueBox(params$confirmatory_cut_point, "Confirmatory Cut Point", icon = "fa-check")
```

### Confirmatory
```{r}
samples %>%
  filter(True_Positive) %>%
  tally() %>%
  pull %>%
gauge(., min = 0, max = nrow(samples), label = "True Positives")
```



Row {data-width=650}
-----------------------------------------------------------------------

### Drug vs No-Drug Signals

```{r}
samples %>%
  mjs_plot(x = Signal_Response_Drug, y = Signal_Response_No_Drug) %>%
  mjs_point(color_accessor = True_Positive) %>%
  mjs_labs("Signal Response Drug", "Signal Response No Drug") %>%
  mjs_add_legend("color_accessor")
```

Row 
-----------------------------------------------------------------------

### True Positive - Drug Signals

```{r}
samples %>%
  filter(True_Positive) %>%
  mjs_plot(x = Signal_Response_Drug) %>%
  mjs_histogram()
```

### True Positive - No Drug Signals

```{r}
samples %>%
  filter(True_Positive) %>%
  mjs_plot(x = Signal_Response_No_Drug) %>%
  mjs_histogram()
```

### True Positive - Signal Inhbition

```{r}
samples %>%
  filter(True_Positive) %>%
  mjs_plot(x = Percent_Signal_Inhibition_Drug) %>%
  mjs_histogram()
```