summary_covid_dbs_samples.Rmd

---
title: "Covid19 DBS Olink sample summary"
author: "Creator: Tea Dodig-Crnkovic"
date: "`r format(Sys.time(), '%d %B, %Y')`"
output:
  html_document:
    toc: true
    code_folding: hide
    toc_float:
      collapsed: false 

---
<!-- highlight: textmate -->

<!-- Edit TOC style (background color) -->
<style>
.list-group-item.active, .list-group-item.active:focus, .list-group-item.active:hover {
    background-color: #EAEAEA;
    color: black;}
</style>

<!-- Set background color for highlighted text -->
<style>
div.highlight { 
  background-color: #EAEAEA;  <!-- #39AC91 (turquoise) -->
  color: #FFFFEF;
  padding: 10px;
  border-radius: 2px 2px 2px 2px;}
</style>

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE, tidy=FALSE)
```


<style type="text/css">

h1.title {
  font-size: 22px;
}
h1 { /* Header 1 */
  font-size: 18px;
}
h2 { /* Header 2 */
    font-size: 15px;
}
h3 { /* Header 3 */
  font-size: 12px;
}

</style>

***
# Description

<div class = "highlight">
Summary of samples included in Covid19 DBS Olink panels CVD3, MET and CAM.   

* Frequency table, demographics and meta data   
  
 
</div>
***

# Initialization
```{r initialization, eval=T, warning=F}
# empty global environment
rm(list = ls(all.names = TRUE))
```

# Load packages
```{r packages, eval=T, warning=F, message=FALSE}
# load packages
library(tidyverse)
library(DT)
library(table1)
```

***
# Load files
```{r pathways, eval=t}
source("load_cov_npx.R")
cov_npx <- load_cov_npx("abspqn_excl")
data_set <- cov_npx$data_set
data_file_path <- cov_npx$data_path
```

Loaded data set **`r cov_npx$data_set`**  
from **`r cov_npx$data_path`**, with the following treatment:  
`r cov_npx$norm`

***

# Select data
```{r data_selection, eval=t}

# select input data
dat <- cov_npx

sinfo  <- as.data.frame(dat$sinfo)
binder <- as.data.frame(dat$binder)
npx <- as.matrix(dat$npx)
norm_info <- cov_npx$norm

# select samples and binders to include
selected_samples <- sinfo$sample_id
selected_binders <- binder$unique_id

samp_info <- paste0("Samples n=", dim(npx[selected_samples, selected_binders])[1])
binder_info <- paste0("Binders n=", dim(npx[selected_samples, selected_binders])[2])

```

```{r}
ds3 <- load_cov_npx(paste0("ds3_", cov_npx$data_set))
sinfo_ds3 <- ds3$sinfo %>% rename(Analysis_group = grp) 
npx_ds3 <- ds3$npx
sinfo_ds3 <- sinfo_ds3 %>% filter(sample_id %in% npx_ds3$sample_id)
binder_ds3 <- ds3$binfo
```

***
**Data options**

*Data set*:   
`r cov_npx$data_set`

*Data set input file*:   
`r cov_npx$data_path`

*Data set notes*:   
`r norm_info`

*Number of samples included:*   
`r samp_info`

*Number of binders included:*   
`r binder_info`

***
# Frequency tables (demographics)
```{r fig.align='center', warning=F}

# only inlcude unique population participant samples (remove duplicates)
dat <- sinfo %>% 
	filter(grp == "Population" | grp == "Bridge") %>% 
	filter(!c(grp == "Bridge" & Analysis_date == "2020-11-13")) %>% # only keep bridge info from first Olink run
  select(Sex,
         Age_grp,
         Symptom,
         Breath,
         Positive,
         serostatus,
         unique_participant_ID,
         umap_seropositive_IgG_IgM,
         Analysis_date,
         Population_batch,
  			 Analysis_group) %>% 
  distinct() %>%
  mutate(serostatus = case_when(serostatus == "no reactivity" ~ "Seronegative, IgM-IgG-",
                                serostatus == "IgG + IgM" ~ "Seropositive, IgM+IgG+",
                                serostatus == "IgM" ~ "Early phase, IgM+IgG-",
                                serostatus == "IgG" ~ "Late phase, IgM-IgG+"))
	
# dat$p_val <- NA

dat$Analysis_group <- factor(dat$Analysis_group)

dat <- 
	dat %>% as_tibble()


label(dat$Sex) <- "Sex"
label(dat$Age_grp) <- "Age group"
label(dat$Symptom) <- "Symptom"
label(dat$Breath) <- "Breath"
label(dat$Positive) <- "Positive"
label(dat$serostatus) <- "Seropositivity group"
label(dat$umap_seropositive_IgG_IgM) <- "UMAP seropositive"
label(dat$Analysis_date) <- "Analysis date in lab"
label(dat$Population_batch) <- "Population batch"
label(dat$Analysis_group) <- "Analysis group"
# label(dat$p_val) <- "P value"

rndr <- function(x, name, ...) {
  if (length(x) == 0) {
    y <- tg[[name]]
    s <- rep("", length(render.default(x=y, name=name, ...)))
    p <- fisher.test(table(y, droplevels(tg$serostatus), useNA = "no"))$p.value
    
    s[2] <- sub("<", "&lt;", format.pval(p, digits=3, eps=0.001))
    s
  } else {
    render.default(x=x, name=name, ...)
  }
}
rndr.strat <- function(label, n, ...) {
    ifelse(n==0, label, render.strat.default(label, n, ...))
}
```


## UMAP group
```{r fig.align='center', warning=F}

# UMAP analysis group
tg <- dat %>% 
	filter(Analysis_group == "UMAP_1" | Analysis_group == "UMAP_0") %>%
  mutate(serostatus = factor(serostatus, levels = c(sort(unique(.$serostatus)), "P-value")))

table1(~ Sex + 
			 	 Age_grp +
         Symptom +
         Breath +
         Population_batch | serostatus, data=tg,
       droplevels=F, render= rndr, render.strat=rndr.strat, overall=F)

```

## SPK group
```{r fig.align='center', warning=F}

# SPK analysis group
tg <- dat %>%
  filter(str_detect(Analysis_group, "SPK")) %>%
  mutate(serostatus = factor(serostatus, levels = c(sort(unique(.$serostatus)), "P-value")))

table1(~ Sex + 
			 	 Age_grp +
         Symptom +
         Breath +
         Population_batch | serostatus, data=tg,
       droplevels=F, render= rndr, render.strat=rndr.strat, overall=F)

```

## Dataset 3

```{r}
# Make table for the table1 table
tg <- sinfo_ds3 %>%
  filter(!str_detect(sample_id, "_[:alpha:]")) %>%
  select(sex, age_group, symptoms, Analysis_group) %>%
  mutate(Analysis_group = case_when(Analysis_group == "Negative" ~ "Seronegative, IgG-",
                                    Analysis_group == "Positive" ~ "Seropositive, IgG+")) %>%
  mutate(Analysis_group = factor(Analysis_group, levels = c(
    sort(unique(as.character(Analysis_group))), "P-value"
  ))) %>%
  rename(serostatus = Analysis_group)

label(tg$sex) <- "Sex"
label(tg$age_group) <- "Age group"
label(tg$symptoms) <- "Symptoms"

table1(~ sex + age_group + symptoms | serostatus, data = tg,
       droplevels = F, render = rndr, render.strat = rndr.strat, overall = F)
```


## UMAP vs SPK group
```{r fig.align='center', warning=F}

# UMAP vs SPK analysis group
tg <- dat %>% 
	mutate(Analysis_date = factor(Analysis_date, levels = c("2020-06-23", "2020-11-13","P-value")))


rndr <- function(x, name, ...) {
  if (length(x) == 0) {
    y <- tg[[name]]
    s <- rep("", length(render.default(x=y, name=name, ...)))
    p <- fisher.test(table(y, droplevels(tg$Analysis_date), useNA = "no"))$p.value
    
    s[2] <- sub("<", "&lt;", format.pval(p, digits=3, eps=0.001))
    s
  } else {
    render.default(x=x, name=name, ...)
  }
}
rndr.strat <- function(label, n, ...) {
    ifelse(n==0, label, render.strat.default(label, n, ...))
}


table1(~ Sex + 
			 	 Age_grp +
         Symptom +
         Breath +
			 	 Analysis_group +
			 	 serostatus +
         Population_batch | Analysis_date, data=tg,
       droplevels=F, render= rndr, render.strat=rndr.strat, overall=F,
			 footnote = "All population participants with their single sample included here")

```
***


# Session information
```{r session_info, eval=T, echo=F}
sessionInfo()
```