Create Multi-omics data integration.Rmd

vignettes/Multi-omics data integration.Rmd

@@ -0,0 +1,235 @@
+---
+title: "Multi-omics Data Integration"
+author:
+  - name: Christina Schmidt
+    affiliation:
+      - Heidelberg University
+  - name: Macabe Daley
+    affiliation:
+      - Heidelberg University
+output:
+  html_document:
+    self_contained: true
+    toc: true
+    toc_float: true
+    toc_depth: 5
+    code_folding: show
+vignette: >
+  %\VignetteIndexEntry{Standard Metabolomics}
+  %\VignetteEncoding{UTF-8}
+  %\VignetteEngine{knitr::rmarkdown}
+bibliography: bibliography.bib
+editor_options:
+  chunk_output_type: console
+  markdown:
+    wrap: sentence
+---
+
+```{=html}
+<style>
+.vscroll-plot {
+    width: 850px;
+    height: 500px;
+    overflow-y: scroll;
+    overflow-x: hidden;
+}
+</style>
+```
+```{r chunk_setup, include = FALSE}
+knitr::opts_chunk$set(
+    collapse = TRUE,
+    comment = "#>"
+)
+```
+
+# <img src="Hexagon_MetaProViz.png" align="right" width="200"/>
+
+\
+[In this tutorial we showcase how to use **MetaProViz** prior knowledge and integrate metabolomics with proteomics:]{style="text-decoration:underline"}:\
+- 1. Load example data and enhance metabolite IDs present.\
+- 2. Use MetaLinksDB to build a metabolite-receptor and metabolite-transporter network.\
+- 3. Use Gaude gene-metabolite sets to perform gene-metabolite enrichment analysis.\
+
+\
+First if you have not done yet, install the required dependencies and load the libraries:
+
+```{r message=FALSE, warning=FALSE}
+# 1. Install Rtools if you haven’t done this yet, using the appropriate version (e.g.windows or macOS).
+# 2. Install the latest development version from GitHub using devtools
+#devtools::install_github("https://github.com/saezlab/MetaProViz")
+
+library(MetaProViz)
+
+library(stringr)
+
+```
+
+\
+\
+
+::: {.progress .progress-striped .active}
+::: {.progress-bar .progress-bar-success style="width: 100%"}
+:::
+:::
+
+# 1. Loading the example data
+
+::: {.progress .progress-striped .active}
+::: {.progress-bar .progress-bar-success style="width: 100%"}
+:::
+:::
+
+\
+[As part of the **MetaProViz** package you can load the example data using the function `toy_data()`]{style="text-decoration:underline"}:\
+For this vignette we will focus on ccRCC patients tissue data:\
+\
+## Metabolomics:
+Here we chose publicly available data from the [paper](https://www.cell.com/cancer-cell/comments/S1535-6108(15)00468-7#supplementaryMaterial) "An Integrated Metabolic Atlas of Clear Cell Renal Cell Carcinoma", which includes metabolomic profiling on 138 matched clear cell renal cell carcinoma (ccRCC)/normal tissue pairs. We have performed differential analysis (details can be found in the vignette [Metadata Analysis](https://saezlab.github.io/MetaProViz/articles/Metadata%20Analysis.html)) and here we load the differential metabolite analysis results for the comparison of Tumour versus Normal.\
+```{r}
+### Metabolomics:
+# Load the example data:
+Metab_TvsN <- MetaProViz::ToyData(Data="Tissue_DMA")
+```
+\
+
+```{r}
+# Add additional potential IDs:
+Metab_TvsN <- MetaProViz::EquivalentIDs(InputData= Metab_TvsN,
+                                        SettingsInfo = c(InputID="Group_HMDB"),# ID in the measured data, here we use the HMDB ID
+                                        From = "hmdb")
+
+```
+
+--> Christina/Macabe: add additional IDs, maybe translate from KEGG to HMDB and from pubchem to HMDB. 
+--> Check what happened to pubchem IDs! --> can also be used for mapping!
+
+
+
+## Proteomics:
+--> Christina: explain sircle and link to appers .etc
+
+```{r}
+### Proteomics:
+Prot_TvN <- MetaProViz::ToyData(Data="Tissue_TvN_Proteomics")
+
+```
+
+::: {.progress .progress-striped .active}
+::: {.progress-bar .progress-bar-success style="width: 100%"}
+:::
+:::
+# 2. MetaLinksDB (metabolite-receptor & metabolite-transporter sets)
+
+::: {.progress .progress-striped .active}
+::: {.progress-bar .progress-bar-success style="width: 100%"}
+:::
+:::
+
+## 2.1 Load and contextualize MetaLinksDB
+The MetaLinks database is a manually curated database of metabolite-receptor and metabolite-transporter sets that can be used to study the connection of metabolites and receptors or transporters [@Farr_Dimitrov2024].\
+To remove potential false positives and decrease the number of putative metabolite-receptor interactions, we filter the MetalinksDB resource to metabolites that are annotated as present in the kidney, blood, or urine in HMDB and known to be extracellular.\
+```{r}
+# Selection as described in ST2 of Farr_Dimitrov2024:
+MetaLinksDB_Res <- MetaProViz::LoadMetalinks(cell_location =c("Extracellular"), 
+                                             tissue_location = c("Kidney", "All Tissues"),
+                                             biospecimen_location = c("Blood",  "Urine"))
+
+MetaLinksDB <- MetaLinksDB_Res[["MetalinksDB"]]
+```
+
+## 2.2. MetaLinksDB coverage in measured data
+First we merge the measured data with our contextualised prior knowledge:\
+```{r}
+# Add metabolomics data
+MetaLinksDB <- merge(x= MetaLinksDB,
+                     y= Metab_TvsN%>% dplyr::rename_with(~ paste0(.x, "_Metab")),
+                     by.x="hmdb",
+                     by.y="Group_HMDB_Metab",
+                     all.x=TRUE)
+
+# Add proteomics data
+MetaLinksDB <- merge(x= MetaLinksDB,
+                     y= Prot_TvN %>% dplyr::rename_with(~ paste0(.x, "_Prot")),
+                     by.x="gene_symbol",
+                     by.y="gene_name_Prot",
+                     all.x=TRUE)
+
+
+# Filter
+MetaLinksDB_Select <- MetaLinksDB%>%
+  dplyr::filter(!is.na(t.val_Metab) | !is.na(t.val_Prot))#only keep MetaLinksDB entries that contain one of the two datatypes.
+  
+```
+
+--> Macabe/Christina: Adapt merge for any possible ID using MetaProViz::CheckMatchID()
+--> Macabe: Just plot and add reference to the prior knwoledge vignette for long explanation
+
+## 2.3. Enrichment analysis
+To perform enrichment analysis, we joined the differential results of proteomics and metabolomics with the metabolite-receptor interactions from MetalinksDB and calculated the mean of the t-values to obtain differential abundance summaries for each interaction and correct for multiple testing using
+the false discovery rate.
+```{r}
+# Calculate mean t-values
+MetaLinksDB_Select$Mean_tval <- (ifelse(is.na(MetaLinksDB_Select$t.val_Metab), 0, MetaLinksDB_Select$t.val_Metab) + # Set NAs to 0 to also include cases where we do not detect a pair?
+                                 ifelse(is.na(MetaLinksDB_Select$t.val_Prot), 0, MetaLinksDB_Select$t.val_Prot)) / 2
+```
+
+
+
+--> Macabe/Christina: Same method as in original publication
+
+## 2.4. Visualisation
+--> Macabe: we can plot everything or the selection based on enrichment analysis scores. Or label top scored pairs as you did with MOFA results (at least if thats easy in R networkplots). 
+
+
+::: {.progress .progress-striped .active}
+::: {.progress-bar .progress-bar-success style="width: 100%"}
+:::
+:::
+# 3. Gene-Metabolite Sets
+
+::: {.progress .progress-striped .active}
+::: {.progress-bar .progress-bar-success style="width: 100%"}
+:::
+:::
+
+## 3.1 Load and convert Gaude gene-sets to gene-metabolite set
+Here we load the Gaude [@Gaude2016] gene-set and convert the gene names to metabolite names using a PK network of metabolic reactions calls CosmosR [@Dugourd2021].\
+With this, we can perform combined pathway enrichment analysis on metabolite-gene sets, if you have other data types such as proteomics/transcriptomics measuring the enzymes expression.\
+
+```{r}
+#Load the example gene-sets:
+MetaProViz::LoadGaude()
+
+#Translate gene names to metabolite names
+Gaude_GeneMetab <- MetaProViz::Make_GeneMetabSet(Input_GeneSet=Gaude_Pathways,
+                                                     SettingsInfo=c(Target="gene"),
+                                                     PKName="Gaude")
+
+Gaude_GeneMetabSet <- Gaude_GeneMetab[["GeneMetabSet"]] 
+```
+
+## 3.2. Gaude coverage in measured data
+--> Macabe: Just plot: Here maybe select all data covered and make a volcano plot of each data type (Proteomics X proteins of X, metabolomics)
+
+
+## 3.3. Gene-Metabolite Set Enrichment analysis
+ORA
+
+## 3.4. Visualisation
+
+--> Macabe: I dont think network makes sense here, but volcano plots of each pathway, having shapes for uniqueness and colour for protein/metabolites
+
+
+::: {.progress .progress-striped .active}
+::: {.progress-bar .progress-bar-success style="width: 100%"}
+:::
+:::
+
+# Session information
+
+```{r session_info, echo=FALSE}
+options(width = 120)
+sessionInfo()
+```
+
+# Bibliography