README.md

Getting Started

# INSTALLATION
git clone https://github.com/conchoecia/genome_assembly_pipelines
cd genome_assembly_pipelines && make

# EXAMPLE RUN 
# We want to insert small scaffolds into chromosomes with Hi-C data.
#  First we must copy the example config file to the directory in which we want to run the program.
cp genome_assembly_pipelines/example_configs/config_GAP_sort_scaffolds_by_hic.yaml ./config.yaml
# then we run snakemake using the associated Snakefile. All the files are saved in the current directory.
snakemake --snakefile genome_assembly_pipelines/scripts/GAP_sort_scaffolds_by_hic_insert

Table of Contents

• Getting Started
• Users' Guide
  • Installation
  • General usage
  • Use cases
    • Insert small scaffolds into chromosome-scale scaffolds
  • Getting help
  • Citing GAP

Users' Guide

This is a collection of snakemake pipelines that are designed to aid in the de novo assembly of genomes.

Movitvation: As much as we would like genome assembly to be a one-command operation, it involves many steps, parameters to optimize, and comparisons between assembly versions. These tasks become more complicated when working with highly heterozygous species.

Currently these scripts are largely undocumeted and they are 100% unpublished. Documentation will be added below as needed for collaborations or publications.

Installation

GAP was implemented in a linux environment with a bash shell, python 3, and snakemake. Every pipeline is implemented in snakemake, and additional required programs are installed locally when executing make or at runtime.

The only steps for installation are to clone the repository and make the included programs.

git clone https://github.com/conchoecia/genome_assembly_pipelines
cd genome_assembly_pipelines && make

General usage

The general workflow for GAP requires the user to copy an example configuration file into the directory in which they wish the program to be run. I recommend doing this in an empty directory to keep the analysis organized and to avoid overwriting and existing config.yaml file.

cp genome_assembly_pipelines/example_configs/config_GAP_sort_scaffolds_by_hic.yaml ./config.yaml

You will find instructions for how to fill out the configuration file inside the configuration file itself.

vim config.yaml

After setting up the configuration file, snakemake is used to run the pipeline. You should refer to the snakemake documentation to determine the number of cores to run, or how to configure snakemake for a cluster environment like SLURM. Here, we add the options -r -p to have more informative messages be printed to the terminal.

snakemake -r -p --snakefile genome_assembly_pipelines/scripts/GAP_sort_scaffolds_by_hic_insert

Output files are typically contained within a single folder of the same name as the tool that was run.

Use cases

These scripts all pertain to various operations used in genome assembly.

Insert small scaffolds into chromosome-scale scaffolds

If you have a chromosome-scale genome assembly with many small scaffolds, it is desirable to insert those small scaffolds into the chromosomes if there is some evidence to do so. Hi-C reads provide such evidence.

This script maps Hi-C data to the genome assembly, and uses a rolling window to determine the location in the chromsomes to which the scaffold has the strongest Hi-C interaction signal. The user selects what quantile of Hi-C interaction strength, and what minimum scaffold size they wish to insert. The program applies these cutoffs as a logical AND.

Insertions into the chromosomes are made only by adding the small scaffodls between existing contigs. No contigs are broken in this procedure. The input .fasta file and the output .fasta file have the same number of contigs.

After the insertion process, there now remains small scaffolds that are not inserted into the chromosome-scale scaffolds, but still have Hi-C connections to the chromosome-scale scaffolds. These are sorted in the fasta file based on their strongest Hi-C location in the chromosome. These are all appended to the .fasta file after the chromosome-scale scaffolds.

Finally, there may be scaffolds that have no Hi-C connections to the chromosome-scale scaffolds. These are appended to the end of the .fasta file in no particular order.

Consideration 1: If you run this program before running purge_dups or purge_haplotigs, then it is important to manually curate the genome assembly and remove duplicate sequences based on the Hi-C map. If there is community demand for a more explicit explanation, please add an issue in the github issues tab above.

Consideration 2: This program does not replace manual genome curation, it only makes this step easier by removing the intial steps of inserting scaffolds.

Getting help

This is currently the only documentation for these scripts. For problems that arise please use the github issues tab above.

Citing GAP

There currently is no publication for citing the GAP package. Please consider citing the github repository directly.