app.qmd

---
title: "doubletroubledb"
format: 
  dashboard:
      logo: images/doubletroubledb_logo.png
      favicon: images/doubletroubledb_logo.png
include-in-header:
  - text: |
      <link rel = "shortcut icon" href = "images/doubletroubledb_logo.png" />
---

```{r}
#| context: setup
#| message: false
#| warning: false
library(ggplot2)
library(tidyverse)
library(ggtree)

# Load data
load(here::here("data", "metadata_all.rda"))
load(here::here("data", "trees.rda"))
load(here::here("data", "dup_counts.rda"))
load(here::here("data", "urls.rda"))

# Load functions
source(here::here("R", "visualization.R"))
```

# Explore

Here, you can explore the relative contribution of each duplication mode
to the duplicated gene repertoire of all species in instances of 
Ensembl (release 110) and Ensembl Genomes (release 57). 

Duplication modes
include segmental (**SD**), tandem (**TD**), proximal (**PD**), 
retrotransposed (**rTRD**), transposed (**TRD**), and dispersed (**DD**) 
duplications.

## Row {.tabset}

### Plants

```{r}
#| width: 30%

# Plot tree - Plants
tree_plants <- plot_tree(
    tree = trees$Plants, 
    metadata = metadata_all$Plants, 
    taxon = "order"
) +
    ggtree::theme_tree2() + 
    theme(
        axis.ticks = element_blank(), 
        axis.line.x = element_blank(), 
        axis.text.x = element_text(color = "white"),
        legend.position = "none",
        plot.margin = margin(0, 0, 0, 0, "cm")
    )

tp <- plotly::ggplotly(tree_plants, tooltip = "label")
for(t in seq_along(tp$x$data)) {
    tp$x$data[[t]]$text <- gsub("tooltip: ", "", tp$x$data[[t]]$text)
}

tp
```

```{r}
#| width: 70%

# Plot barplot - Plants
bar_plants <- dup_counts$Plants |>
    dplyr::mutate(
        species = factor(species, levels = rev(get_taxa_name(tree_plants)))
    ) |>
    group_by(species) |>
    mutate(
        perc = n / sum(n),
        perc = round(perc * 100, 2)
    ) |>
    ungroup() |>
    mutate(
        text = paste0(
            "Species: ", species, "\n",
            "Type: ", type, "\n",
            "N: ", n, "\n",
            "%: ", perc
        )
    ) |>
    plot_duplicate_freqs(plot_type = "stack_percent") +
    theme(
        axis.text.y = element_blank(),
        axis.ticks.y = element_blank(),
        plot.margin = margin(0, 0, 0, 0, "cm")
    ) +
    coord_cartesian(expand = FALSE) +
    labs(y = NULL)

bp <- plotly::ggplotly(bar_plants, tooltip = "text")
for(t in seq_along(bp$x$data)) {
    bp$x$data[[t]]$text <- gsub("text: ", "", bp$x$data[[t]]$text)
}

bp
```

### Vertebrates

```{r}
#| width: 30%

# Plot tree - Vertebrates
tree_vert <- plot_tree(
    tree = trees$Vertebrates, 
    metadata = metadata_all$Vertebrates, 
    taxon = "class"
) +
    ggtree::theme_tree2() + 
    theme(
        axis.ticks = element_blank(), 
        axis.line.x = element_blank(), 
        axis.text.x = element_text(color = "white"),
        legend.position = "none",
        plot.margin = margin(0, 0, 0, 0, "cm")
    )

tv <- plotly::ggplotly(tree_vert, tooltip = "label")
for(t in seq_along(tv$x$data)) {
    tv$x$data[[t]]$text <- gsub("tooltip: ", "", tv$x$data[[t]]$text)
}

tv
```

```{r}
#| width: 70%

bar_vert <- dup_counts$Vertebrates |>
    dplyr::mutate(
        species = factor(species, levels = rev(get_taxa_name(tree_vert)))
    ) |>
    group_by(species) |>
    mutate(
        perc = n / sum(n),
        perc = round(perc * 100, 2)
    ) |>
    ungroup() |>
    mutate(
        text = paste0(
            "Species: ", species, "\n",
            "Type: ", type, "\n",
            "N: ", n, "\n",
            "%: ", perc
        )
    ) |>
    plot_duplicate_freqs(plot_type = "stack_percent") +
    theme(
        axis.text.y = element_blank(),
        axis.ticks.y = element_blank(),
        plot.margin = margin(0, 0, 0, 0, "cm")
    ) +
    coord_cartesian(expand = FALSE) +
    labs(y = NULL)

bv <- plotly::ggplotly(bar_vert, tooltip = "text")
for(t in seq_along(bv$x$data)) {
    bv$x$data[[t]]$text <- gsub("text: ", "", bv$x$data[[t]]$text)
}

bv
```

### Metazoa

```{r}
#| width: 30%

# Plot tree - Vertebrates
tree_mz <- plot_tree(
    tree = trees$Metazoa, 
    metadata = metadata_all$Metazoa, 
    taxon = "phylum"
) +
    ggtree::theme_tree2() + 
    theme(
        axis.ticks = element_blank(), 
        axis.line.x = element_blank(), 
        axis.text.x = element_text(color = "white"),
        legend.position = "none",
        plot.margin = margin(0, 0, 0, 0, "cm")
    )

tm <- plotly::ggplotly(tree_mz, tooltip = "label")
for(t in seq_along(tm$x$data)) {
    tm$x$data[[t]]$text <- gsub("tooltip: ", "", tm$x$data[[t]]$text)
}

tm
```

```{r}
#| width: 70%

# Create barplot
bar_mz <- dup_counts$Metazoa |>
    dplyr::mutate(
        species = factor(species, levels = rev(get_taxa_name(tree_mz)))
    ) |>
    group_by(species) |>
    mutate(
        perc = n / sum(n),
        perc = round(perc * 100, 2)
    ) |>
    ungroup() |>
    mutate(
        text = paste0(
            "Species: ", species, "\n",
            "Type: ", type, "\n",
            "N: ", n, "\n",
            "%: ", perc
        )
    ) |>
    plot_duplicate_freqs(plot_type = "stack_percent") +
    theme(
        axis.text.y = element_blank(),
        axis.ticks.y = element_blank(),
        plot.margin = margin(0, 0, 0, 0, "cm")
    ) +
    coord_cartesian(expand = FALSE) +
    labs(y = NULL)

bmz <- plotly::ggplotly(bar_mz, tooltip = "text")
for(t in seq_along(bmz$x$data)) {
    bmz$x$data[[t]]$text <- gsub("text: ", "", bmz$x$data[[t]]$text)
}

bmz
```


### Fungi

```{r}
#| width: 30%

# Plot tree - Fungi
tree_f <- plot_tree(
    tree = trees$Fungi, 
    metadata = metadata_all$Fungi, 
    taxon = "phylum"
) +
    ggtree::theme_tree2() + 
    theme(
        axis.ticks = element_blank(), 
        axis.line.x = element_blank(), 
        axis.text.x = element_text(color = "white"),
        legend.position = "none",
        plot.margin = margin(0, 0, 0, 0, "cm")
    )

tf <- plotly::ggplotly(tree_f, tooltip = "label")
for(t in seq_along(tf$x$data)) {
    tf$x$data[[t]]$text <- gsub("tooltip: ", "", tf$x$data[[t]]$text)
}

tf
```

```{r}
#| width: 70%

# Create barplot
bar_f <- dup_counts$Fungi |>
    dplyr::mutate(
        species = factor(species, levels = rev(get_taxa_name(tree_f)))
    ) |>
    group_by(species) |>
    mutate(
        perc = n / sum(n),
        perc = round(perc * 100, 2)
    ) |>
    ungroup() |>
    mutate(
        text = paste0(
            "Species: ", species, "\n",
            "Type: ", type, "\n",
            "N: ", n, "\n",
            "%: ", perc
        )
    ) |>
    plot_duplicate_freqs(plot_type = "stack_percent") +
    theme(
        axis.text.y = element_blank(),
        axis.ticks.y = element_blank(),
        plot.margin = margin(0, 0, 0, 0, "cm")
    ) +
    coord_cartesian(expand = FALSE) +
    labs(y = NULL)

bf <- plotly::ggplotly(bar_f, tooltip = "text")
for(t in seq_along(bf$x$data)) {
    bf$x$data[[t]]$text <- gsub("text: ", "", bf$x$data[[t]]$text)
}

bf
```

### Protists

```{r}
#| width: 30%

# Plot tree - Fungi
tree_p <- plot_tree(
    tree = trees$Protists, 
    metadata = metadata_all$Protists, 
    taxon = "phylum"
) +
    ggtree::theme_tree2() + 
    theme(
        axis.ticks = element_blank(), 
        axis.line.x = element_blank(), 
        axis.text.x = element_text(color = "white"),
        legend.position = "none",
        plot.margin = margin(0, 0, 0, 0, "cm")
    )

tp <- plotly::ggplotly(tree_p, tooltip = "label")
for(t in seq_along(tp$x$data)) {
    tp$x$data[[t]]$text <- gsub("tooltip: ", "", tp$x$data[[t]]$text)
}

tp
```

```{r}
#| width: 70%

# Create barplot
bar_p <- dup_counts$Protists |>
    dplyr::mutate(
        species = factor(species, levels = rev(get_taxa_name(tree_p)))
    ) |>
    group_by(species) |>
    mutate(
        perc = n / sum(n),
        perc = round(perc * 100, 2)
    ) |>
    ungroup() |>
    mutate(
        text = paste0(
            "Species: ", species, "\n",
            "Type: ", type, "\n",
            "N: ", n, "\n",
            "%: ", perc
        )
    ) |>
    plot_duplicate_freqs(plot_type = "stack_percent") +
    theme(
        axis.text.y = element_blank(),
        axis.ticks.y = element_blank(),
        plot.margin = margin(0, 0, 0, 0, "cm")
    ) +
    coord_cartesian(expand = FALSE) +
    labs(y = NULL)

bp <- plotly::ggplotly(bar_p, tooltip = "text")
for(t in seq_along(bp$x$data)) {
    bp$x$data[[t]]$text <- gsub("text: ", "", bp$x$data[[t]]$text)
}

bp
```

# Download

Use the column filters and search bar to find your species of interest. Then,
click the species name to download a list of all duplicate pairs
and duplicated genes.

```{r}
# Reduce list of data frames to a single data frame
dmetadata <- dplyr::bind_rows(metadata_all, .id = "Ensembl") |>
    inner_join(urls, by = "species") |>
    mutate(
        species_name = paste0(
            "<a href='", url, "' target='_blank'>", ncbi_species, "</a>"
        )
    )

# Keep only useful columns
cols <- c(
    Species = "species_name", TaxID = "taxonomy_id",
    Assembly = "assembly", Accession = "assembly_accession",
    Family = "family", Order = "order", Class = "class", 
    Phylum = "phylum"
)
final_table <- dmetadata[, cols]
names(final_table) <- names(cols)
final_table$TaxID <- as.character(final_table$TaxID)


# Create DataTable
dt <- DT::datatable(
    final_table,
    selection = 'single',
    rownames = FALSE,
    filter = 'top',
    escape = FALSE,
    options = list(
        lengthMenu = c(5, 10, 25, 50, 100), 
        pageLength = 10
    )
)

dt
```

# FAQ

::: {.callout-note icon="false"}

## 1) How do I cite this resource?

To cite `doubletroubledb` in publications, use:

> Almeida-Silva, F., & de Peer, Y. V. (2024). doubletrouble: an R/Bioconductor package for the identification, classification, and analysis of gene and genome duplications. BioRxiv. https://doi.org/10.1101/2024.02.27.582236


A BibTeX entry for LaTeX users is:

```
 @article{Almeida-Silva2024.02.27.582236,
	author = {Fabricio Almeida-Silva and Yves Van de Peer},
	title = {doubletrouble: an R/Bioconductor package for the identification, classification, and analysis of gene and genome duplications},
	year = {2024},
	doi = {10.1101/2024.02.27.582236},
	publisher = {Cold Spring Harbor Laboratory},
	URL = {https://www.biorxiv.org/content/early/2024/02/29/2024.02.27.582236},
	journal = {bioRxiv}
}
```

:::

---

::: {.callout-note icon="false"}

## 2) How do I report an issue or suggest a feature?

You can open an issue in the [GitHub repository](https://github.com/almeidasilvaf/doubletroubledb) 
where the source code for this app is.

:::

---

::: {.callout-note icon="false"}

## 3) Where can I find the code used to create the data in this resource?

All data in this resource (duplicate pairs, duplicated genes, 
and .collinearity files) were generated for a benchmark in 
[the paper associated with this app](https://doi.org/10.1101/2024.02.27.582236).
All code used in this paper are available in the pages below:

1. GitHub repository: <https://github.com/almeidasilvaf/doubletrouble_paper>.
2. Quarto book: <https://almeidasilvaf.github.io/doubletrouble_paper/>.

:::

---

::: {.callout-note icon="false"}

## 4) The tree topology is not 100% accurate in a specific species tree. Why is that?

Since Ensembl doesn't provide species trees for all its instances, we
inferred a species tree using a standard approach in phylogenomics.
The following fragment was extracted from the manuscript associated
with this publication:

> BUSCO scores (Manni et al. 2021) for each species were obtained and 
visualized using the Bioconductor package 
cogeqc (Almeida‐Silva and Van de Peer 2023). BUSCO genes shared by >90% of 
the species were aligned with MAFFT (Katoh and Standley 2013), and 
multiple sequence alignments were concatenated and trimmed to remove 
alignment columns with >50% gaps. Filtered supermatrices were used for 
phylogeny inference with IQ-TREE2 (Minh et al. 2020). Oomycetes, red algae, 
*Giardia lamblia*, *Mnemiopsis leidyi*, and
*Saccharomyces cerevisiae* were used as outgroups for Ensembl Fungi, Ensembl
Plants, Ensembl Protists, Ensembl Metazoa, and Ensembl, respectively.


Although inferring trees from BUSCO genes is common practice, this method
has limitations and can lead to wrong topologies. Our goal here was not
test hypotheses on the phylogeny of each Ensembl instance (with more
sophisticated methods such as partition models, microsynteny-based phylogenies,
and gene tree-species tree reconciliation methods, for instance). Instead,
we simply wanted to have species trees that could be used as a phylogenetic
context to explore major patterns in the frequency of duplicated genes
by mode.


If you have a curated species tree that includes all species in one (or more)
of the Ensembl instances used here, we'd be happy to update the tree in this
app. You can contribute data by opening an issue in
[this GitHub repo](https://github.com/almeidasilvaf/doubletroubledb).

:::