Improving our assemblies

Since a couple of years, many shortcomings of our assemblies became clearer. Once of the biggest issues concerns spatial information, mainly caused by the huge number of scaffolds and no idea about their actual position along chromosomes. Because of that we'll try to use a chromosome-level assembly to reorder our assemblies accordingly. Given that we have no clue about accuracy, we will try three different mappers: minimap2, mummer and lastal.

Before starting

1. Unzip JGI assembly with gzip -d assemblies/Alyrata_384_v1_only1-8.fa.gz
2. Add A. halleri and A. lyrata assemblies in assemblies folder. You can download them here and here. Please rename the two assemblies to Ahal_v2_2.fa and Alyr_v2_2.fa. If you prefer other names, please modify all scripts in scripts/ accordingly
3. Install Conda or miniconda3 and run conda create env -f envs/mappers.yaml -n mappers to create a Conda environment with (almost) everything we need.
4. Clone the RagTag github repository: git clone https://github.com/malonge/RagTag.git
5. Activate mappers via conda activate mappers and install RagTag via python setup.py install (within the RagTag folder)

To start

First activate mappers (if you didn't already) via conda activate mappers, then to run all analyses please stay on the RemappingAssemblies/ folder and use:

sh scripts/minimap.sh
sh scripts/mummer.sh
sh scripts/lastal.sh

Results

In the results folder, for each method and each assembly, there is a txt file with different scores for each contig of our assembly. Another agp file provides information about the inferred location in the JGI assembly.