oncotator

oncotator

Install oncotator

Anaconda3 python3

conda create -n oncotator python=2.7
source activate oncotator
pip install numpy
pip install pyvcf
git clone https://github.com/broadinstitute/oncotator.git
cd oncotator
python setup.py install
pip install biopython or pandas .....

run oncotator

oncotator.py [-h] [-v] -i file --input-format Format
            [--output-format Format] [--qsub node]

1 输入为VCF格式

2 输入为MAFLITE格式

NCBI_Build	sample	chr	start	end	ref_allele	alt_allele
hg19	7	55259515	55259515	T	G
hg19	7	140453136	140453136	A	T
hg19	1	120612003	120612004	GG	-
hg19	8	145138175	145138176	-	G
hg19	12	42538391	42538401	GGAGCGAGCAG	-

Oncotator Error解决方法记录

1. 注释 varscan VCF出错

由于VCF格式不正确(oncotator只能识别标准VCF格式)

	标准VCF	varScan VCF
##头部注释	无RD	有RD
##头部注释	##FORMAT=<ID=AD,Number=R	##FORMAT=<ID=AD,Number=1
FORMAT: AD	复合形式 :RD,AD:	分别显示 RD:AD

解决方法

1. 修改头部注释 ##FORMAT=<ID=AD,Number=1 改为 ##FORMAT=<ID=AD,Number=R, 否则出现 ValueError: invalid literal for float(): 932,1
2. 修改FORMAT AD的部分 改为与标准VCF格式一致，AD包括ref_depth,alt_depth 代码如下

awk -F '\t' '$0 !~ /^##FORMAT=<ID=RD/ {OFS="\t";
if($0 ~ /^##FORMAT=<ID=AD/){gsub("Number=1", "Number=R",$0); print $0; next};
if($0 ~ /^#/){print $0; next};
split($9, a, ":"); split($10, b, ":");
for(i in a){if(a[i] == "RD"){ind = i}}; gsub("RD:", "", $9);
b[ind + 1] = b[ind]","b[ind + 1]; delete b[ind]; c = b[1];
for(i = 2; i <= length(b); i++){if(i != ind){c = c":"b[i]}};
$10 = c; print $0}' varScan.vcf > NEW.varScan.vcf