dianhe/outcrossseq

OutcrossSeq

OutcrossSeq is a software package for genotyping and imputation in the recombinant populations of outcrossing plants, based on whole-genome low-coverage resequencing data.

Availability

You can download the source code and the Demo data here:
https://github.com/xhhuanglab/OutcrossSeq You can download the Docker container here:

docker run -it -v $PWD:/docker mjchenjojo/outcrossseq:v1.0
#$PWD is the path of your data and perl scripts, you need to put the script and the input vcf file in a folder

Dependencies

The code is implemented in perl (v5.16.3) and can be run from linux shell. To run OutcrossSeq the following libraries and tools should be available:

1. R version 3.4.1
2. Statistics::Basic module
3. Number::Format module

Software structure

There are three modules in OutcrossSeq.

• "Diploid-Outcrossing" module designed for diploid outcrossing F₁ population
• "Double-Cross" module designed for double cross population
• "Autopolyploid Plant" module designed for autopolyploid F₁ population.

Data preparation

The input file of "Diploid-Outcrossing" module is variant calling file of two heterozygous parents and all F₁ individuals (in VCF format). See Demo data: demo_diploid.vcf

The input file of "Double-Cross" module is variant calling file of four inbred parents and all offspring individuals. See Demo data:demo_double.vcf

The input files of "Autopolyploid Plant" module are two separated variant calling files for parents and offsprings, respectively. For parents, variants are called in polyploid mode, for offsprings variants are called in diploid mode. See Demo data for two hexaploid parents: demo-Parents.vcf and their F₁ offsprings demo_autopolyploid.vcf

• For variant calling，we recommend to generate VCF files for each chromosome, which is more convenient to run the code and to tune parameters in following steps.

Getting Started

For details, please see Use cases. "Diploid-Outcrossing" module

cd Diploid-Outcrossing

perl system_diploid_cross.pl 5 100 0.6 0.3 0.35 60 1000 100000 parent1 parent2 demo_diploid

Parameter	Description
5	chromosome
100	quality threshold we set here as an example, and SNP sites with quality below threshold will be filtered
0.6	miss rate threshold we set here as an example, and missing values greater than the threshold will be filtered
0.3	heterozygosity between 0.3 and (1-0.3)
0.35	start cuttree threshold
60	sample number
1000	min marker number in each window, at least 1000
100000	window size (<1cM)
parent1	the id of parent1 in vcf file
parent2	the id of parent2 in vcf file
demo_diploid	the input file of the vcf file

perl diploid_cross_step4.pl 5 5 demo

Parameter	Description
first 5	start chromosome
second 5	end chromosome
demo	output file's name

"Double-Cross" module

cd Double-Cross

perl system_double_cross.pl 16 17 18 19 5 100 100000 1 6 demo_double

Parameter	Description
16 17 18 19	column number of four parents;16—parent1,17—parent2,18—parent3,19—parent4(Caution: 16,17 should be inbred parent1 x parent2 ,the column 18,19 should be inbred parent3 x parent4. If mix up, the result will be wrong)
5	chromosome
100	quality threshold we set here as an example, and SNP sites with quality below threshold will be filtered
100000	window size (<1cM)
1	the first sample order
6	the last sample order
demo_double	the input file of the vcf file e

perl double_cross_step3.pl 5 5 6 100000

Parameter	Description
first 5	star of chromosome
second 5	end of chromosome
6	Size of the population
100000	window size (<1cM)

"Autopolyploid Plant" module

cd Autopolyploid-Plant

perl system_autopolyploid.pl 5 0.08 1 50 0.9 3 100 Parent1 Parent2 demo_autopolyploid-Parents demo_autopolyploid

Parameter	Description
5	chromosome
0.08, 1	depth range from 0.08 to 1
50	min SNP marker number of each window
0.9	cuttree
3	select min cuttree marker num
100	quality threshold we set here as an example, and SNP sites with quality below threshold will be filtered
Parent1	the id of parent1 in vcf file)
Parent2	the id of parent2 in vcf file
demo_autopolyploid-Parents	the input file of the parents vcf file
demo_autopolyploid	the input file of the offsprings vcf file

perl autopolyploid_step5.pl 5 5

Parameter	Description
first 5	start chromosome
second 5	end chromosome

Table of Contents

• Users' Guide
  • Installation
  • Introduction
  • Use cases
    • Diploid-Outcrossing
    • Double-Cross
    • Autopolyploid Plant
  • Getting help
  • Citing OutcrossSeq

Users' Guide

Installation

git clone https://github.com/xhhuanglab/OutcrossSeq.git
cd OutcrossSeq
#list three modules
ls 
#autopolyploid_crop  diploid_outcrossing_species  double_cross_populations  readme.md 

Introduction

OutcrossSeq has three modules: "Diploid-Outcrossing", "Double-Cross" and "Autopolyploid Plant". Typically the work includes three steps – sequencing, genotype calling and imputation.

Step1 sequencing: Using Tn5 enzymes for low-cost library construction, hundreds of individuals are then pooled within the same library (each individual with its unique barcode). The pools are sequenced using second-generation sequencing technology and further demultiplexed according to the index tag.

Step2 genotype calling: GATK can be used for genotype calling from whole-genome low-coverage sequencing data.

Step3 imputation: we have three different strategies for different species ("Diploid-Outcrossing" module, "Double-Cross" module and “Autopolyploid Plant” module).

Use cases

OutcrossSeq works with VCF format files as input. After imputation, the output file is a genotype file that could be directly used for QTL mapping.

"Diploid-Outcrossing" and "Double-Cross" outputs are in bin matrix format, which can be used as the inputs of GACD, MapQTL etc for genetic mapping.

“Autopolyploid Plant” outputs are in SNP matrix format, which can be used as the inputs of fastGWA, EMMAX etc for genetic mapping.

For more details, please check Demo run below.

Diploid-Outcrossing module

This module is designed for all diploid outcrossing species. F1 population with about 200 individuals is good enough for a well study using this module.

1:Filter out low-quality SNP sites

cd /OutcrossSeq/diploid_outcrossing_species
perl diploid_cross_Genotype_filter.pl 5 parent1 parent2 100 0.6 demo_diploid 0.3

Parameter	Description
5	chromosome
parent1	the id of parent1 in vcf file
parent2	the id of parent2 in vcf file
100	quality threshold we set here as an example, and SNP sites with quality below threshold will be filtered
0.6	miss rate threshold we set here as an example, and missing values greater than the threshold will be filtered
demo_diploid	the input file of the vcf file
0.3	heterozygosity between 0.3 and (1-0.3)

2:Calculate kinship of each sample in each windows

perl diploid_cross_step1.pl 100000 1000 F1-chr5.geno 5 0.35

Parameter	Description
100000	window size (<1cM)
1000	min marker number in each window, at least 1000
F1-chr5.geno	the input file
5	chromosome
0.35	start cuttree threshold

3:Clustering in each window

perl diploid_cross_step2.pl 5 60 0.35

Parameter	Description
5	chromosome
60	sample number
0.35	start cuttree threshold

Start cuttree threshold can be determine by the evolutionary tree, as example in the figure,you can set a threshold > 0.3. You can randomly select several windows on the chromosome to draw the evolutionary tree to determine the threshold.

4:Haplotying across the chromosome

In this step, a sample.list is required which records the order of individuals in VCF file.

#less sample.list
Sim1
Sim2
Sim3
Sim4
Sim5
Sim6
Sim7
Sim8
Sim9
Sim10
...

perl diploid_cross_step3.pl 5

Parameter	Description
5	chromosome

5:Generate the genotype file

perl diploid_cross_step4.pl 5 5 demo

Parameter	Description
5	chromosome
demo	output file's name

Optional steps: One may want to check the haplotypes of parents using following command line

perl diploid_cross_Haplotype.pl demo.geno demo_diploid parent1 parent2

Parameter	Description
demo.geno	the genotype of bin map
demo_diploid	the file of vcf file such as demo_diploid.vcf
parent1	the id of parent1 in vcf file
parent2	the id of parent2 in vcf file

Double-Cross module

This module is designed for typical Double-Cross population in crop breeding programs, such as rice and maize etc. Population size with 500 individuals is sufficient to carry out a nice study. It can also be applied to other types of populations when the founders are no more than four inbred lines.

1.filter out low-quality low-rate variants and select homozygous variants without missing data and filter the variants that two parents of hybridization have the same mutation and then transform vcf file to genotype file

cd /OutcrossSeq/double_cross_populations
perl double_cross_Genotype_filter.pl  demo_double 100 16 17 18 19

Parameter	Description
demo_double	the input file of the vcf file
100	quality threshold we set here as an example, and SNP sites with quality below threshold will be filtered
16 17 18 19	column number of four parents;16—parent1,17—parent2,18—parent3,19—parent4(Caution: 16,17 should be inbred parent1 x parent2 ,the column 18,19 should be inbred parent3 x parent4. If mix up, the result will be wrong)

2:Determine the genotype of the first allele

perl double_cross_step1.pl demo_double_parent.geno demo_double_offspring.geno  100000 5 1 6

Parameter	Description
demo_double_parent.geno,demo_double_offspring.geno	input file
100000	window size (<1cM)
5	chromosome
1	the first sample order
6	the last sample order

3:Determine the genotype of the second allele

perl double_cross_step2.pl demo_double_parent.geno demo_double_offspring.geno  100000 5 1 6

Parameter	Description
demo_double_parent.geno,demo_double_offspring.geno	input file
100000	window size (<1cM)
5	chromosome
1	the first sample order
6	the last sample order

4:Generate the genotype file In this step, a sample.list is required which records the order of individuals in VCF file.

#less sample.list
Sim1
Sim2
Sim3
Sim4
Sim5
Sim6

perl double_cross_step3.pl 5 5 6 100000

Parameter	Description
5	star of chromosome
5	end of chromosome
6	Size of the population
100000	window size (<1cM)

Autopolyploid Plant module

This module is designed dedicated for F₁ population of autopolyploid or near-autopolyploid plants such as potato and sweet potato etc. A F_{1</sub Population with about 1000 or more individuals is recommended, for a reasonable practice.}

1:Select simplex or double-simplex SNPs

perl autopolyploid_Genotype_filter_step1.pl demo_autopolyploid-Parents 100 0.08 1

Parameter	Description
demo_autopolyploid-Parents	the input file of the parents vcf file
100	quality threshold we set here as an example, and SNP sites with quality below threshold will be filtered
0.08,1	depth range from 0.08 to 1

2:Transform .vcf to .geno file

perl autopolyploid_Genotype_filter_step2.pl 5 demo_autopolyploid Parent1 Parent2

Parameter	Description
5	chromosome
demo_autopolyploid	the input file of the offsprings vcf file
Parent1	the id of parent1 in vcf file
Parent2	the id of parent2 in vcf file

3:Filter low-quality and error ratio sites This step can be skipped

perl autopolyploid_step1.pl 5 demo_autopolyploid

Parameter	Description
5	chromosome
demo_autopolyploid	the input file of the .geno file

4:Select relevant sites

perl autopolyploid_step2.pl select-demo_autopolyploid 5 50 0.9 3 100

Parameter	Description
select-demo_autopolyploid	the input file of the select*.geno file
5	chromosome
50	min SNP marker number of each window
0.9	cuttree
3	select min cuttree marker num
100	select max cuttree marker num

Optimized steps: Then one has to combine all the results together when all runs finished

perl split-select-geno.pl 5 110 50 select-demo_autopolyploid

Parameter	Description
5	chromosome
110	File lines (Can be obtained by command:wc -l file_name) of select-demo_autopolyploid.geno
50	line number of plit file,be careful this parameter must be a multiple of the min snp number for each window
select-demo_autopolyploid	the split file name

Optimized steps: Then one has to combine all the results together when all runs finished

perl combine_split.pl 5 select-demo_autopolyploid 2

Parameter	Description
5	chromosome
select-demo_autopolyploid	the split file name
3	Number of split files

*5:Select mark readsNum * In this step, a sample.list is required which records the order of individuals in VCF file.

#less sample.list
Sim1
Sim2
Sim3
Sim4
Sim5
Sim6
Sim7
Sim8
Sim9
Sim10
...

perl autopolyploid_step3.pl demo_autopolyploid-readsNum.txt R2-select-demo_autopolyploid

Parameter	Description
demo_autopolyploid-readsNum.txt	input file1
R2-select-demo_autopolyploid	input file2 named R2-select-demo_autopolyploid.txt

6:Imputation missing SNPs

perl autopolyploid_step4.pl 5 R2-select-demo_autopolyploid-readsNum.txt R2-select-demo_autopolyploid.txt

Parameter	Description
demo_autopolyploid-readsNum.txt	input file1
R2-select-demo_autopolyploid	input file2

7:Merge genotype files

perl autopolyploid_step5.pl 5 5
#because demo data only contains the chromosome 5

Parameter	Description
first 5	start chromosome
second 5	end chromosome

Getting help

Citing OutcrossSeq

The paper describing OutcrossSeq can be found here: