主要是参考这个进行的，https://github.com/humanlongevity/HLA 其文章在这：http://www.pnas.org/content/early/2017/06/27/1707945114
首先是拿到fastq文件，然后进行了下fastqc质控，发现质量不咋地，死马当活马医，试试。
然后是bwa比对，先是
bwa index /home/biolinux/reference/hg38.fa #建立索引
bwa mem /home/biolinux/reference/hg38.fa /home/biolinux/Downloads/HLA.fastq > hla.sam #然后比对
samtools view -S hla.sam -b > hla.bam #格式转换
samtools sort hla.bam hla_sorted.bam # 排序
samtools index hla_sorted.bam # 索引
最后，进行xhla算法比对。
docker run -v `pwd`:`pwd` -w `pwd` humanlongevity/hla --sample_id test --input_bam_path hla.bam --output_path test
得出结果，一个比较疑惑的地方是hg38(without alt contigs)没搞懂。有待解决。
{
"subject_id": "test",
"creation_time": "2017-09-19T07:09:32Z",
"report_version": "1.2",
"report_type": "hla_typing",
"sample_id": "test",
"hla": {
"alleles": [
"A*24:290",
"A*24:290",
"C*07:02",
"C*07:02",
"DPB1*13:01",
"DPB1*33:01",
"DQB1*06:11",
"DQB1*06:39",
"DRB1*15:01",
"DRB1*16:01"
]
}
}
