在如何在IGV上使用BLAT搜索非模式物种中,我们讨论如何用Apache的CGI服务器,来响应IGV发送的BLAT请求。考虑到我们大部分的时候都是个人使用,并不需要Apache这种重量级的Web服务器,因此完全可以省去这一组件。
我之前学过Python和Django和Flask这两种网页应用框架,其中Flask比较轻量,非常适合我们这种小型应用。因此,我就用Flask编写了一个Blat网页应用,用来响应IGV的请求。
首先,安装Flask(为了避免冲突,我用了虚拟环境)
mkdir myproject
cd myproject
python3 -m venv venv
# 启动
. venv/bin/activate
pip3 install flask
接着,从我的GitHub上下载脚本,见https://github.com/xuzhougeng/myscripts/blob/master/misc/myBlat.py
或者复制下面的代码进行修改,命名为myBlast.py
from flask import Flask
from flask import request
import tempfile
import subprocess
from os.path import basename
app = Flask(__name__)
"""
注意:
1. 修改ref_dict中的ref_name1和ref_name2fa的位置
2. 修改query_seq_db中的 ref1 和 ref2, 改成你的物种里的特有关键词, 例如A.thaliana 我用的是thaliana
"""
ref_dict = {
"ref_name1": "/path/to/your/reference1.fa",
"ref_name2": "/path/to/your/reference2.fa"
}
def query_seq_db(db):
#if "\" in db:
# dbname = db.split("\")[-1]
#else:
# dbname = basename(db)
#dbname = dbname.split(".")[0]
#ref = ref_dict[dbname]
#return ref
db = db.lower()
if "ref1" in db:
return ref_dict["ref_name1"]
elif "ref2" in db:
return ref_dict["ref_name2"]
else:
return ref_dict["ref_name1"]
def blat(seq,db,type):
# 需要改成你实际的blat路径
blat = "/opt/biosoft/ucsctools/blat/blat"
# 需要改成你实际的参考位置
ref = query_seq_db(db)
fd1, fa_file = tempfile.mkstemp(suffix=".fa")
fd2, psl_file = tempfile.mkstemp(suffix=".psl")
with open(fa_file,"w") as f:
f.write(f">query\n{seq}\n")
process = subprocess.Popen([blat, ref, fa_file, psl_file],
stdout=subprocess.PIPE,
stderr=subprocess.PIPE)
stdout,stderr = process.communicate()
psl_ret = [line for line in open(psl_file, "r") ]
psl_rec = [line.strip().split('\t') for line in psl_ret[5:] ]
return psl_rec
@app.route("/myBlat", methods=['GET'])
def myBlat():
seq = request.args.get('userSeq','')
db = request.args.get('db','')
output = request.args.get('output','json')
type = request.args.get('type', 'DNA')
psl_rec = blat(seq, db, type)
data = {
"track": "blat",
"genome": "ath",
"fields": ["matches", "misMatches", "repMatches", "nCount", "qNumInsert", "qBaseInsert", \
"tNumInsert", "tBaseInsert", "strand", "qName", "qSize", "qStart", "qEnd", "tName", \
"tSize", "tEnd", "blockCount", "blockSizes", "qStarts", "tStarts"],
"blat" : psl_rec
}
return data
```
最后运行Flask
```nginx
flask run -p 5000
```
需要注意的是,我用的端口是5000,而不是80,因此还需要在IGV偏好中BLAT URL改成
```nginx
http://127.0.0.1:5000/myBlat?userSeq=$SEQUENCE&type=DNA&db=$DB&output=json
```