---
title: "6. Tutorial: Download histology images from the Human Protein Atlas"
author:
- name: Anh N. Tran
  affiliation: DataGrata LLC
  email: trannhatanh89@gmail.com
date: "`r Sys.Date()`"
output: 
    BiocStyle::html_document:
        toc: true
        toc_depth: 2
        toc_float: true
        number_sections: true
vignette: >
  %\VignetteIndexEntry{"6. Tutorial: Download histology images from the Human Protein Atlas"}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(
  collapse=TRUE,
  comment="#>",
  warning=FALSE,
  error=FALSE,
  eval=FALSE
)
```

```{r library, message=FALSE, warning=FALSE, error=FALSE}
library(BiocStyle)
library(HPAanalyze)
library(dplyr)
```

# The case 
What to do if you want to systematically download images from HPA?

# The solution

## Get the download links

```{r}
CCNB1xml <- hpaXmlGet("ENSG00000134057")

CCNB1_ab <- hpaXmlAntibody(CCNB1xml)
CCNB1_ab
#> # A tibble: 4 x 4
#>   id        releaseDate releaseVersion RRID      
#>   <chr>     <chr>       <chr>          <chr>     
#> 1 CAB000115 2006-03-13  1.2            <NA>      
#> 2 CAB003804 2006-10-30  2              AB_562272 
#> 3 HPA030741 2013-12-05  12             AB_2673586
#> 4 HPA061448 2016-12-04  16             AB_2684522

CCNB1_expr <- hpaXmlTissueExpr(CCNB1xml)
str(CCNB1_expr[[1]])
#> Classes 'tbl_df', 'tbl' and 'data.frame':    331 obs. of  18 variables:
#>  $ patientId         : chr  "1653" "1721" "1725" "598" ...
#>  $ age               : chr  "53" "60" "57" "7" ...
#>  $ sex               : chr  "Male" "Female" "Male" "Male" ...
#>  $ staining          : chr  NA NA NA NA ...
#>  $ intensity         : chr  NA NA NA NA ...
#>  $ quantity          : chr  NA NA NA NA ...
#>  $ location          : chr  NA NA NA NA ...
#>  $ imageUrl          : chr  "http://v18.proteinatlas.org/images/115/2043_B_4_5.jpg" "http://v18.proteinatlas.org/images/115/2043_B_6_5.jpg" "http://v18.proteinatlas.org/images/115/2043_B_5_5.jpg" "http://v18.proteinatlas.org/images/115/2043_A_2_2.jpg" ...
#>  $ snomedCode1       : chr  "M-00100" "M-00100" "M-00100" "M-00100" ...
#>  $ snomedCode2       : chr  "T-93000" "T-93000" "T-93000" "T-66000" ...
#>  $ snomedCode3       : chr  NA NA NA NA ...
#>  $ snomedCode4       : chr  NA NA NA NA ...
#>  $ snomedCode5       : chr  NA NA NA NA ...
#>  $ tissueDescription1: chr  "Normal tissue, NOS" "Normal tissue, NOS" "Normal tissue, NOS" "Normal tissue, NOS" ...
#>  $ tissueDescription2: chr  "Adrenal gland" "Adrenal gland" "Adrenal gland" "Appendix" ...
#>  $ tissueDescription3: chr  NA NA NA NA ...
#>  $ tissueDescription4: chr  NA NA NA NA ...
#>  $ tissueDescription5: chr  NA NA NA NA ...
```

## Download the images

```{r eval=FALSE}
dir.create("img")
for (i in 1:nrow(CCNB1_expr[[1]])) {
    download.file(CCNB1_expr[[1]]$imageUrl[i],
                  destfile = paste0("img/", CCNB1_ab$id[1], "_",
                                    CCNB1_expr[[1]]$patientId[i], "_",
                                    CCNB1_expr[[1]]$tissueDescription2[i],
                                    ## the extra i below ensures unique file name
                                    i, ".jpg"),
                  mode = "wb")
}
```

## Notes

While the HPA website displays two images for many of the samples, this method only provides one. 

If you wish to filter the results by tissueDescription or snomedCode, it's important to note that the value you are searching for may be present in any of the columns. To address this, you can use the `filter_all(any_vars())` function from the `dplyr` package.

# Copyright

__Anh Tran, 2018-2023 __

Please cite: __Tran, A.N., Dussaq, A.M., Kennell, T. et al. HPAanalyze: an R package that facilitates the retrieval and analysis of the Human Protein Atlas data. BMC Bioinformatics 20, 463 (2019) https://doi.org/10.1186/s12859-019-3059-z __