06-Genome.Rmd

```{r setupgenome, include=FALSE}
rm(list = ls()) ; invisible(gc()) ; set.seed(42)
library(knitr)
library(tidyverse)
library(raster)
library(leaflet)
library(sf)
theme_set(bayesplot::theme_default())
opts_chunk$set(
  echo = F, message = F, warning = F, fig.height = 6, fig.width = 8,
  cache = T, cache.lazy = F)
```


# Genome assembly

*Assembly of high-quality annotated reference genomes for three trees of three different species.*

## Sequencing

* CNRGV, Toulouse
* Genomic reads with PacBio HiFi
* Optical maps with Bionano
* Illumina HiSeq/NovaSeq

## Assembly workflow

### V1

The first assembly will be produced by William Marande in the CNRGV.

### V2

A second version can be produced using a `singularity` and `snakemake` workflow developed by Ludovic Duvaux (possible collaboration).
The wrokflow uses the following steps:

1. converting `bam` files to `fastq`: converting HiFi reads to regular fastq with the package `simlink`
1 assembly of contigs: using `hifiasm` first but step to test additional assembler such as `canu` and `spades`
1. removing haplotigs: [`purge_dups`](https://github.com/dfguan/purge_dups) and `purge_haplotigs`
1. optical maps: proprietary program of bionano & manual curation by CNRGV to obatin a haploid optical map
1. gap closing: using Illumina reads with [`SOAPdenovo2`](https://sourceforge.net/projects/soapdenovo2/files/GapCloser/bin/r6/)
1. quality check:
    * [`Jellyfish`](https://github.com/gmarcais/Jellyfish)
    * [KAT plot](https://www.researchgate.net/publication/309161336_The_present_and_future_of_de_novo_whole-genome_assembly)
    * BUSCO
    * synteny:
        * Dgeanies
        * jupiterPlot
    * quast

## Annotation workflow

Automatic singularity & snakemake workflow to annotate genomes: https://github.com/sylvainschmitt/genomeAnnotation .

## Project description

> In each species, using a single genotype, such as leaves from a single branch, we will extract DNA and obtain long reads -2 runs of each Mini and Prometh ION targeting a total coverage of 100x -and short reads -sequencing one of the libraries to be used for the detection of mutations (below) using Illumina NovaSeq6000  technology  and  targeting  a  minimum  coverage  of  100x.  The  sequencing  services  will  be subcontracted to CEA Genoscope, Evry. Ideally, we would use cambium tissue to produce the reference genomes, however, since a limited amount of cambium tissue can be sampled per tree, we will use leaf tissue. Assembling the quality-filtered reads from both sequencing technologies will provide long scaffolds (Johnson et al., 2019). In each species, we will build an “optimal map” of the genome using high-resolution restriction   maps   from   single,   labelled   molecules   of   DNA   obtained   through   Bionano   technology (subcontracted  to  CNRGV,  Toulouse).  In  combination  with  the  previous  scaffolds  this  will provide  a chromosomal assembly.The three genomes will be annotated using automated pipelines (Bolger, Arsova, & Usadel, 2018).

> Seedling original genotype (T4.1): Short read sequencing data will be demultiplexed, quality-filtered, and mapped to the reference genomes. The “zero mutation reference” genotype will be constructed from the consensus genotype  calls  obtained  from  short  read  sequencing  of  the  three  cambium  samples  using  a standard variant calling approach such as  GATK (“Best Practices for Variant Calling with the GATK,” 2015).