---
title: "23-Apul-miRNA-methylation-machinery"
author: "Kathleen Durkin"
date: "2025-01-27"
always_allow_html: true
output: 
  github_document:
    toc: true
    toc_depth: 3
    number_sections: true
    html_preview: true 
  bookdown::html_document2:
    theme: cosmo
    toc: true
    toc_float: true
    number_sections: true
    code_folding: show
    code_download: true
editor_options: 
  markdown: 
    wrap: 72
bibliography: ../../references.bib
---

```{r}
library(dplyr)
```

One way miRNAs and DNA methylation can interact to influence expression is through miRNAs influencing the transcription/translation of protein machinery involved in DNA methylation. In humans, this has been identified as a potential mechanism of carcinogenisis! (Reviewed in @karimzadeh_regulation_2021)

## Which genes are involved in DNA methylation?

"DNA methylation machinery is composed of several enzymes to accurately regulate the gene expression together. The main enzyme-encoding genes comprise a small family of DNA cytosine-5 methyltransferase, which are involved in established (DNA methyltransferase 3 alpha (DNMT3A) and DNA methyltransferase 3 beta (DNMT3B)) and maintenance (DNMT1) of methylation patterns. The other genes are methyl-CpG-binding proteins (MBPs; methyl-CpG-binding domain (MBD)1–4, methyl-CpG-binding protein 2 (MECP2), UHRF1/2, ZBTB33, ZBTB4, and ZBTB38) that assist in expression repression and 10–11 translocation (TET) enzymes (TET1, TET2, and TET3); seemingly, are implicated in DNA demethylation [10, 11]. Some experiments demonstrated that, based on the importance of oncogenes and/or tumor suppressor genes alterations in the development of each type of cancer, DNA methylation regulator genes may enhance or suppress cancer development in each specific type of cancer [12,13,14,15,16]."
@karimzadeh_regulation_2021

So we're looking for miRNAs that target genes encoding:
- DNA methyltransferases (DNMTs)
- methyl-CpG-binding proteins
- 10–11 translocation (TET) enzymes

First let's see what a quick search of our annotated A.pulchra genome turns up. I want to find any annotations related to DNA methyltransferase

```{r, engine='bash'}

awk -F'\t' '$5 ~ /methyltransferase/ {print $2, $5}' OFS='\t' \
../output/02-Apul-reference-annotation/Apulcra-genome-mRNA-IDmapping-2024_12_12.tab \
| grep "DNA"

```

There's definitely DNMT machinery genes anontated in our A.pulchra genome (as expected)! Now let's create a data frame that associated each annotated mRNA with it's gene ID (i.e. FUN######) and do a more broad search for different machinery involved in methylation

```{r}
# Read in data
mRNA_annot <- read.table("../output/02-Apul-reference-annotation/Apulcra-genome-mRNA-IDmapping-2024_12_12.tab", header=TRUE, sep='\t') %>% select(-X)
mRNA_geneIDs <- read.table("../output/15-Apul-annotate-UTRs/Apul-mRNA-FUNids.txt", header=FALSE, sep='\t')

# Join
mRNA_annot_geneID <- left_join(mRNA_annot, mRNA_geneIDs, by="V1")
mRNA_annot_geneID$V4 <- gsub("Parent=", "", mRNA_annot_geneID$V4)

# To double-check that this worked correctly, check for NAs in the column containing gene IDs
# Every mRNA has a gene ID, so this should return FALSE
any(is.na(mRNA_annot_geneID$V3.y))
```

Now let's select genes that encode different protein machinery relevant to DNA methylation

```{r}
methyl_machinery <- subset(mRNA_annot_geneID, 
                           (grepl("methyltransferase", mRNA_annot_geneID$Protein.names) 
                            & grepl("DNA", mRNA_annot_geneID$Protein.names))
                           | grepl("DNMT", mRNA_annot_geneID$Protein.names)
                           | grepl("Dnmt", mRNA_annot_geneID$Protein.names)
                           | grepl("MBP", mRNA_annot_geneID$Protein.names)
                           | grepl("MBD", mRNA_annot_geneID$Protein.names)
                           | grepl("MECP", mRNA_annot_geneID$Protein.names)
                           | grepl("UHRF", mRNA_annot_geneID$Protein.names)
                           | grepl("ZBTB", mRNA_annot_geneID$Protein.names)
                           | grepl("TET", mRNA_annot_geneID$Protein.names)
                           )

nrow(methyl_machinery)

```

We find 23 mRNAs in our A.pulchra genome that encode known DNA methylation machinery!

Note that I stipulated the term "methyltransferase" must be accompanied by "DNA" also present in the Protein.names column. I imposed this restriction because, when I searched for "methyltransferase" alone, methyltransferases related to histone modification and modification with nucleotides other than cytosine came up.

Now let's see whether any of these methylation machinery genes are putative miRNA targets, using the output from our target-prediction software, miRanda.

```{r}
miranda_apul <- read.table("../output/18-Apul-interactions-functional-annotation/miRanda_miRNA_mRNA.txt", header=TRUE, sep='\t')

miranda_methyl_machinery <- left_join(methyl_machinery, miranda_apul, by=c("V4" = "mRNA_FUNid"))

length(na.omit(miranda_apul$miRNA_cluster))
length(unique(na.omit(miranda_apul$miRNA_cluster)))

length(na.omit(miranda_methyl_machinery$miRNA_cluster))
length(unique(na.omit(miranda_methyl_machinery$miRNA_cluster)))
```
Miranda originally identified 6109 putative targets of 39 of our miRNAs, but none of these targets are included in our list of DNA methylation machinery mRNA. :(


So this means that our **A.pulchra miRNAs do not putatively bind to any methylation machinery mRNA** and/or I'm missing some genes in my search.