report_slow.qmd

---
title: "Dereplication and cherry-picking report"
output_format: html
---

```{r pkgs, message=FALSE, echo=FALSE}
library(magrittr)
library(dplyr)
library(knitr)
library(ggplot2)
library(cowplot)
```


## Spectra quality control per-plate

Table with the total number of raw spectra per plate (`n_spectra`), and the total number of valid spectra (`n_valid_specta`), that did not met any of the red flags of quality control:

* the raw spectra has no data (`is_empty`)
* the raw spectra length is different of more than 2 units (`is_outlier_length`)
* the raw spectra was acquired correctly (`is_not_regular`)

```{r qc_plate, echo=FALSE}
targets::tar_read(spectra_stats) %>% 
  dplyr::mutate(maldi_plate = basename(maldi_plate)) %>% 
  dplyr::relocate(maldi_plate) %>% 
  knitr::kable()
```


## Dereplication


```{r , echo=FALSE,results='asis'}
glue::glue(
  "Out of {n_valid_spectra} valid spectra, ",
  "{n_picked} spectra were picked using the {procedure} procedure.",
  n_valid_spectra = nrow(targets::tar_read(picked)),
  n_picked = sum(targets::tar_read(picked)[["to_pick"]]),
  procedure = unique(targets::tar_read(summary_picked)[["procedure"]])
)
```

### Distribution of cluster size

```{r cluster_size, fig.height=6, fig.width=5, fig.dpi=300, echo=FALSE}
targets::tar_read(picked) %>%
  dplyr::filter(!is_edge) %>% 
  dplyr::arrange(cluster_size) %>%
  dplyr::select(membership, cluster_size) %>% unique %>%
  dplyr::mutate(membership = factor(membership, membership)) %>% 
  ggplot(aes(x= membership, y=cluster_size))+
  geom_point()+
  geom_segment(aes(x=membership,xend=membership,y=0,yend=cluster_size))+
  geom_text(aes(label = cluster_size),hjust=-.5, fontface = "bold")+
  scale_y_continuous(expand = expansion(add = c(0, 2)))+
  coord_flip()+
  labs(x="Cluster numeric identifier", y = "# of spectra per cluster")+
  theme_cowplot()
```


### Distribution of OD600 per cluster

```{r od600, fig.height=6, fig.width=5, fig.dpi=300, echo=FALSE}
targets::tar_read(picked) %>%
  dplyr::filter(!is_edge) %>% 
  dplyr::mutate(
    membership = factor(membership),
    membership = forcats::fct_reorder(membership, OD600,max)) %>% 
  ggplot(aes(x= membership, y=OD600, color = to_pick))+
  geom_point(shape = 23, stroke = 1)+
  scale_color_brewer(palette = "Set1", direction = -1,name = "Picked isolate?", labels = c("No", "Yes"))+
  scale_y_continuous(breaks = seq(from =0.5, to = 2.5, by=0.5))+
  coord_flip()+
  labs(x="Cluster numeric identifier",
       title = "OD600 per cluster")+
  theme_minimal_hgrid()
```