基于GATK流程化进行SNP calling

在进行变异检测时，以群体基因组重测序数据为例，涉及到的个体基本都是上百个，而其中大多数流程均是重复的步骤。
本文将基于GATK进行SNP calling的流程写入循环，便于批量分析。
在这里插入图片描述

1 涉及变量

1.工作目录work_dir/
2.参考基因组ref_genome.fa
3.Reads列表read_list.txt
4.测序平台Illumina
5.调用线程数

2 调用数据

1.参考基因组ref_genome.fa
2.重测序数据sample1/sample1_1.fq.gz、sample1/sample1_2.fq.gz……
3.Reads列表：read_list.txt
生成方法：预先将存放各个个体Reads的文件夹放入一个文件夹work_dir/然后使用下列命令生成：

ls work_dir/ > read_list.txt

3 主要脚本

usage:

bash GATK_pipeline.sh work_dir/ ref_genome.fa read_list.txt Illumina 10

`GATK_pipeline.sh`


#---------------------------------------------------------------#
#                objection defined by user                      #
#---------------------------------------------------------------#

set -au

# 1.
# Master dir.:
WORK_dir=$1

# 2.
# Reference genome:
REF=$2

# 3.
# Read list:
READ_list=$3

# 4.
# Seqencing platform:
PL=$4

# 5.
# number of threads:
NT=$5

#---------------------------------------------------------------#
#         main loop for SNPs calling by gatk pipeline           #
#---------------------------------------------------------------#

#READ_list.txt is a list of read groups.
while read -r READ

do

SAMPLE=SM_${READ}
ID=${READ}
READ1="${WORK_dir}${READ}_1.fq"
READ2="${WORK_dir}${READ}_2.fq"
OUT="${READ}"

#1.
#Alignning reads to reference genome by BWA-MEM2-mem, producing a .sam data
bwa-mem2 \
	mem \
	-M \
	-t ${NT} \
	-R "@RG\tID:${ID}\tSM:${SAMPLE}\tPL:${PL}" \
	${REF} \
	${READ1} \
	${READ2} \
	> ${OUT}.sam

#2.
#Sorting .sam by gatk-SortSam, producing a .bam data
gatk \
	SortSam \
	-I ${OUT}.sam \
	-O ${OUT}.bam \
	-SO coordinate \
	-VALIDATION_STRINGENCY LENIENT \
	-CREATE_INDEX true \
	-TMP_DIR ./${OUT}tmp.sort
#3.
#Marking dupulications in .bam by gatk-MarkDuplicates
#producing a .dup.bam and .dup.txt data
gatk \
	MarkDuplicates \
	-I ${OUT}.bam \
	-O ${OUT}.dup.bam \
	-M ${OUT}.dup.txt \
	-REMOVE_DUPLICATES true \
	-VALIDATION_STRINGENCY LENIENT \
	-CREATE_INDEX true \
	-TMP_DIR ${OUT}tmp.dup

#4.
#QC by samtools-flagstat, producing a .dup.bam.stat data
samtools \
	flagstat \
	${OUT}.dup.bam \
	> ${OUT}.dup.bam.stat

#5.
#Calling SNPs by gatk-HaplotypeCaller, producing a .dup.vcf data
gatk \
	HaplotypeCaller \
	-R ${REF} \
	-I ${OUT}.dup.bam \
	-O ${OUT}.dup.vcf

done < $READ_list
##