1、下载安装 https://bitbucket.org/mroachawri/purge_haplotigs/wiki/Install

1、Dependencies (in no particular order)

$ sudo apt install bedtools
$ bedtools --version
bedtools v2.26.0

$ sudo apt install samtools
$ samtools --version
samtools 1.7
Using htslib 1.7-2
Copyright (C) 2018 Genome Research Ltd.

$ sudo apt install r-base r-base-dev

on a new install we wont have the required R library 'ggplot2' installed

$ sudo su - -c "R -e \"install.packages('ggplot2', repos='http://cran.rstudio.com/')\""

# download the latest release from https://github.com/lh3/minimap2/releases (currently v2.13)
$ wget https://github.com/lh3/minimap2/releases/download/v2.13/minimap2-2.13_x64-linux.tar.bz2
$ tar xf minimap2-2.13_x64-linux.tar.bz2

we'll add a bin directory to the home folder and add to the PATH, then install there

$ mkdir ~/bin
$ printf "export PATH=\$PATH:~/bin\n" > .bashrc
$ source .bashrc
$ cp minimap2-2.13_x64-linux/minimap2 ~/bin/

$ minimap2 -V
2.13-r850

# download the latest release from https://github.com/mummer4/mummer/releases (currently 4.0.0.beta2)
$ wget https://github.com/mummer4/mummer/releases/download/v4.0.0beta2/mummer-4.0.0beta2.tar.gz
$ tar xf mummer-4.0.0beta2.tar.gz

compile

$ cd mummer-4.0.0beta2
$ ./configure
$ make
$ cd ../

install (just softlink to the home bin directory ~/bin)

$ ln -s ~/mummer-4.0.0beta2/delta-filter ~/bin/delta-filter
$ ln -s ~/mummer-4.0.0beta2/nucmer ~/bin/nucmer
$ ln -s ~/mummer-4.0.0beta2/show-coords ~/bin/show-coords

$ nucmer -V
4.0.0beta2

2、Install Purge Haplotigs

installing to user's home directory, no compiling, just add the purge_haplotigs/bin directory to the system PATH.

# clone the git
$ git clone https://bitbucket.org/mroachawri/purge_haplotigs.git

create a softlink to ~/bin

$ ln -s ~/purge_haplotigs/bin/purge_haplotigs ~/bin/purge_haplotigs

test Purge Haplotigs

$ purge_haplotigs

USAGE:
purge_haplotigs [options]

COMMANDS:
-- Purge Haplotigs pipeline:
readhist First step, generate a read-depth histogram for the genome
contigcov Second step, get contig coverage stats and flag 'suspect' contigs
purge Third step, identify and reassign haplotigs

-- Other scripts
ncbiplace Generate a placement file for submission to NCBI
test Test everything!

test the pipeline

$ purge_haplotigs test
#
ALL TESTS PASSED

3、Running Purge Haplotigs（https://www.jianshu.com/p/8ed5b494b131）

PREPARATION

minimap2 -t 4 -ax map-pb genome.fa subreads.fasta.gz --secondary=no \
| samtools sort -@ 8 -m 1G -o aligned.bam -T tmp.ali

STEP 1

Generate a coverage histogram by running the first script. This script will produce a histogram png image file for you to look at and a BEDTools 'genomecov' output file that you'll need for STEP 2.

purge_haplotigs hist -b aligned.bam -g genome.fasta [ -t threads ]

STEP 2

Run the second script using the cutoffs from the previous step to analyse the coverage on a contig by contig basis. This script produces a contig coverage stats csv file with suspect contigs flagged for further analysis or removal.

purge_haplotigs cov -i aligned.bam.genecov -l -m -h \
[-o coverage_stats.csv -j 80 -s 80 ]

STEP 3

Run the purging pipeline. This script will automatically run a BEDTools windowed coverage analysis (if generating dotplots), and minimap2 alignments to assess which contigs to reassign and which to keep. The pipeline will make several iterations of purging. Optionally, parse repeats -r in BED format for improved handling of repetitive regions

purge_haplotigs purge -g genome.fasta -c coverage_stats.csv

You will have five files

• .fasta: These are the curated primary contigs
• .haplotigs.fasta: These are all the haplotigs identified in the initial input assembly.
• .artefacts.fasta: These are the very low/high coverage contigs (identified in STEP 2). NOTE: you'll probably have mitochondrial/chloroplast/etc. contigs in here with the assembly junk.
• .reassignments.tsv: These are all the reassignments that were made, as well as the suspect contigs that weren't reassigned.
• .contig_associations.log: This shows the contig "associations" e.g