---
title: "sRNA Comparison"
author: Steven Roberts
date: "`r format(Sys.time(), '%d %B, %Y')`" 
output: 
  github_document:
    toc: true
    toc_depth: 3
    number_sections: true
    html_preview: true
  html_document:
    theme: readable
    highlight: zenburn
    toc: true
    toc_float: true
    number_sections: true
    code_folding: show
    code_download: true

---

```{r setup, include=FALSE}
library(knitr)
library(tidyverse)
library(kableExtra)
library(DESeq2)
library(pheatmap)
library(RColorBrewer)
library(data.table)
library(DT)
library(formattable)
library(Biostrings)
library(spaa)
library(tm)
knitr::opts_chunk$set(
  echo = TRUE,         # Display code chunks
  eval = FALSE,         # Evaluate code chunks
  warning = FALSE,     # Hide warnings
  message = FALSE,     # Hide messages
  fig.width = 6,       # Set plot width in inches
  fig.height = 4,      # Set plot height in inches
  fig.align = "center" # Align plots to the center
)
```

https://gannet.fish.washington.edu/Atumefaciens/20230620-E5_coral-fastqc-flexbar-multiqc-sRNAseq/A_pulchra/trimmed/sRNA-ACR-140-S1-TP2.flexbar_trim.20230621_1.fastq.gz 


```{bash}
wget -r \
--no-directories --no-parent \
-P ../data \
-A "_1.fastq.gz" https://gannet.fish.washington.edu/Atumefaciens/20230620-E5_coral-fastqc-flexbar-multiqc-sRNAseq/A_pulchra/trimmed/
```

```{bash}
gunzip ../data/*gz
```



```{bash}
cat ../data/sRNA-ACR*fastq | awk 'NR%4==2' | wc -l 
```

```{bash}
cat ../data/sRNA-ACR*fastq | awk 'NR%4==2' | sort | uniq | wc -l 
```
```{r}
13018677 / 93798121
```


# evermani

https://gannet.fish.washington.edu/Atumefaciens/20230620-E5_coral-fastqc-flexbar-multiqc-sRNAseq/P_evermanni/trimmed/sRNA-POR-73-S1-TP2.flexbar_trim.20230621_1.fastq.gz


```{bash}
wget -r \
--no-directories --no-parent \
-P ../data \
-A "_1.fastq.gz" https://gannet.fish.washington.edu/Atumefaciens/20230620-E5_coral-fastqc-flexbar-multiqc-sRNAseq/P_evermanni/trimmed/
```

```{bash}
gunzip ../data/sRNA-POR*gz
```



```{bash}
cat ../data/sRNA-POR*fastq | awk 'NR%4==2' | wc -l 
```

```{bash}
cat ../data/sRNA-POR*fastq | awk 'NR%4==2' | sort | uniq | wc -l 
```


Meandrina

https://gannet.fish.washington.edu/Atumefaciens/20230620-E5_coral-fastqc-flexbar-multiqc-sRNAseq/P_meandrina/trimmed/sRNA-POC-48-S1-TP2.flexbar_trim.20230621_1.fastq.gz

```{bash}
wget -r \
--no-directories --no-parent \
-P ../data \
-A "_1.fastq.gz" https://gannet.fish.washington.edu/Atumefaciens/20230620-E5_coral-fastqc-flexbar-multiqc-sRNAseq/P_meandrina/trimmed/
```

```{bash}
gunzip ../data/sRNA-POC*gz
```



```{bash}
cat ../data/sRNA-POC*fastq | awk 'NR%4==2' | wc -l 
```

```{bash}
cat ../data/sRNA-POC*fastq | awk 'NR%4==2' | sort | uniq | wc -l 
```



# blast comparison

```{bash}
cat /home/shared/8TB_HDD_01/sam/data/A_pulchra/sRNAseq/collapsed/*fasta > ../output/apul_coll.fasta
```



```{bash}
cat /home/shared/8TB_HDD_01/sam/data/P_evermanni/sRNAseq/collapsed/*fasta > ../output/peve_coll.fasta
```


```{bash}
cat /home/shared/8TB_HDD_01/sam/data/P_meandrina/sRNAseq/collapsed/*fasta > ../output/pmea_coll.fasta
```

```{r, engine='bash'}
/home/shared/ncbi-blast-2.11.0+/bin/makeblastdb \
-in ../output/apul_coll.fasta \
-dbtype nucl \
-out ../data/blast/apul_sRNA
```

```{r, engine='bash'}
/home/shared/ncbi-blast-2.11.0+/bin/makeblastdb \
-in ../output/peve_coll.fasta \
-dbtype nucl \
-out ../data/blast/peve_sRNA
```

```{r, engine='bash'}
/home/shared/ncbi-blast-2.11.0+/bin/makeblastdb \
-in ../output/pmea_coll.fasta \
-dbtype nucl \
-out ../data/blast/pmea_sRNA
```



```{r, engine='bash', eval=TRUE}
ls ../output/*coll*
```



```{r, engine='bash', eval=TRUE}
ls ../data/blast/*sRNA*
```




```{r, engine='bash'}
/home/shared/ncbi-blast-2.11.0+/bin/blastn \
-query ../output/apul_coll.fasta \
-db ../data/blast/peve_sRNA \
-out ../output/apul_peve_sRNA_blastn.tab \
-evalue 1E-40 \
-num_threads 40 \
-max_target_seqs 1 \
-max_hsps 1 \
-outfmt 6


```

```{r, engine='bash'}
/home/shared/ncbi-blast-2.11.0+/bin/blastn \
-query ../output/apul_coll.fasta \
-db ../data/blast/pmea_sRNA \
-out ../output/apul_pmea_sRNA_blastn.tab \
-evalue 1E-40 \
-num_threads 40 \
-max_target_seqs 1 \
-max_hsps 1 \
-outfmt 6


```

```{r, engine='bash'}
/home/shared/ncbi-blast-2.11.0+/bin/blastn \
-query ../output/peve_coll.fasta \
-db ../data/blast/pmea_sRNAs \
-out ../output/peve_pmea_sRNA_blastn.tab \
-evalue 1E-40 \
-num_threads 40 \
-max_target_seqs 1 \
-max_hsps 1 \
-outfmt 6

```


```{r, engine='bash', eval=TRUE}
grep ">" -c ../output/*coll*
```


```{r, engine='bash', eval=TRUE}
wc -l ../output/*tab
```