ONT_CRISPR_workflow.Rmd

---
title: "ONT CRIPR-Cas9 enrichment workflow"
author: "Marcus Davy, Elena Lopez-Girona"
date: "5/14/2020"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)

library(RLinuxModules)

# module("purge")

module("load openlava/3.2")
module("load asub/2.1")

Sys.setenv(
  INDIR1="/input/genomic/plant/Malus/OxfordNanopore/Red_flesh", 
  WKDIR=file.path("/workspace", Sys.getenv("USER"), "github/analysis-workflows/Malus/Red_Flesh_ON")
)

```


Albacore basecalling and QC
===========================


## Convert one to many fast5 files

```{r engine="bash" }
module load conda/4.8.2
conda activate cflsjt_albacore

mkdir -p $WKDIR/albacore2/Red_flesh_ON_run1_Cas9/single/fasta5/log 

echo "[ albacore multi_to_single_fast5 version ]"
multi_to_single_fast5 -v

if [ ! -e "$WKDIR/albacore2/Red_flesh_ON_run1_Cas9/single/fasta5/1/0007a1c5-4e4c-49f9-a3aa-d3f07aaee6b5.fast5" ]; then
  time multi_to_single_fast5 -i $INDIR1 -s $WKDIR/albacore2/Red_flesh_ON_run1_Cas9/single/fasta5
fi

ls -l $WKDIR/albacore2/Red_flesh_ON_run1_Cas9/single/fasta5
```

_Note: Conda environment does not load using `RLinuxModules` -- use reticulate_


## Base call intensities using albacore2

```{r, engine="bash"}
module load conda/4.8.2
conda activate cflsjt_albacore

THREADS=12
# #mkdir -p $WKDIR/albacore2/Red_flesh_ON_run1_Cas9/${i}
if [ ! -e "$WKDIR/albacore2/Red_flesh_ON_run1_Cas9/0/workspace/pass" ]; then

  for i in $(seq 0 1)
  do
    echo "read_fast5_basecaller.py -f FLO-MIN106 -k SQK-DCS108 -t $THREADS -s $WKDIR/albacore2/Red_flesh_ON_run1_Cas9/${i} \
-o fastq -q 100000 -i $WKDIR/albacore2/Red_flesh_ON_run1_Cas9/single/fasta5/${i}"
  done | asub -n $THREADS -j $WKDIR/albacore2/Red_flesh_ON_run1_Cas9/log

else 

  for i in $(seq 0 1)
  do
    echo "read_fast5_basecaller.py -f FLO-MIN106 -k SQK-DCS108 -t $THREADS -s $WKDIR/albacore2/Red_flesh_ON_run1_Cas9/${i} \
-o fastq -q 100000 -i $WKDIR/albacore2/Red_flesh_ON_run1_Cas9/single/fasta5/${i}"
  done | asub -n $THREADS -j $WKDIR/albacore2/Red_flesh_ON_run1_Cas9/log -p

fi
```

## pycoQC

```{r engine="bash"}
module load conda/4.8.2
conda activate cflsjt_albacore

mkdir -p $WKDIR/00.Fastq
mkdir -p $WKDIR/01.pycoQC/log

cd $WKDIR/albacore2/Red_flesh_ON_run1_Cas9

# for f in $(echo 0 1); do
#  sed '1d' ${f}/sequencing_summary.txt > ${f}_summary.txt
# done

# head -n 1 0/sequencing_summary.txt > header
# cat header *summary.txt > all_summary_run1.txt

cat 0/sequencing_summary.txt       > all_summary_run1.txt
sed '1d' 1/sequencing_summary.txt >> all_summary_run1.txt

bsub -o $WKDIR/01.pycoQC/log/QC1.out -e $WKDIR/01.pycoQC/log/QC1.err -J ONT_QC1 \
  pycoQC -f $WKDIR/albacore2/Red_flesh_ON_run1_Cas9/all_summary_run1.txt -o $WKDIR/01.pycoQC/pycoQC_summary_run1.html
```


## NanoPlot summary

```{r engine="bash"}
## Conflicts with conda environment
module load nanopack/1.0.0

mkdir -p $WKDIR/02.NonoPlot/log

NanoPlot --summary $WKDIR/albacore2/Red_flesh_ON_run1_Cas9/all_summary_run1.txt --loglength -o $WKDIR/albacore2/mkdir -p $WKDIR/02.NonoPlot/summary-plots-log-transformed
```


Guppy basecalling and QC
========================


## Convert one to many fast5 files

```{r engine="bash" }
module load guppy/3.5.2

# mkdir -p $WKDIR/albacore2/Red_flesh_ON_run1_Cas9/single/fasta5/log 

# echo "[ albacore multi_to_single_fast5 version ]"
# multi_to_single_fast5 -v

# if [ ! -e "$WKDIR/albacore2/Red_flesh_ON_run1_Cas9/single/fasta5/1/0007a1c5-4e4c-49f9-a3aa-d3f07aaee6b5.fast5" ]; then
#   time multi_to_single_fast5 -i $INDIR1 -s $WKDIR/albacore2/Red_flesh_ON_run1_Cas9/single/fasta5
# fi

# ls -l $WKDIR/albacore2/Red_flesh_ON_run1_Cas9/single/fasta5

mkdir -p $WKDIR/Guppy_basecalling/log

bsub << EOF
#!/bin/bash
#BSUB -J guppy
#BSUB -o $WKDIR/Guppy_basecalling/log/guppy.out 
#BSUB -e $WKDIR/Guppy_basecalling/log/guppy.err
#BSUB -R "gpu"

guppy_basecaller --input_path $INPUT --save_path $WKDIR/Guppy_basecalling/00.Fastq \
--flowcell FLO-MIN106 --kit SQK-LSK109 -x "cuda:0"

EOF
```


## See also

Albacore base caller;

* https://omictools.com/albacore-tool
* https://github.com/dvera/albacore

Oxford Nanopore;

& https://github.com/nanoporetech/ont_fast5_api


pycoQC;

* https://github.com/a-slide/pycoQC

Tutorials;

* https://pypi.org/project/ont-fast5-api/