Virus Identification On Site

[TOC]

**1.0/directory** includes some ngs softwares, supported with ubuntu16.04. *The image is too fat ,so not recommended.*；

**2.0/derectory** is through ubuntu17.10 repository list including some ngs softwares. *The image is so simple , so recommended.*；

VIOS function is to identify **virus** and confirm its classitication.
From ngs **FASTQ** data, VIOS can achieve that the viruses category 、proportion 、coverage , and so on.

1. set DATABASE variable in settings.py:

   DATABASES = {
       'default': {
           'ENGINE': 'django.db.backends.mysql',
           'NAME': 'vios_db',
           'HOST':'mysqler',  #mysql_container_hostname(--hostname=mysqler)
           'PORT':'3306',
           'USER':'root',
           'PASSWORD':'1234aaaa',
       }
   }
   

2. start myslq container, and appoint the hostname para same with the HOST element of DATABASE variable:

   docker run -it --hostname=mysqler --name mysql -e MYSQL_ROOT_PASSWORD=1234aaaa -e MYSQL_DATABASE=vios_db mysql:5.7
   

3. start vios container, and appoint the link para which can connect to MySQL: create image, first：

   docker image build -t suyanan/vios:1.0 . docker pull suyanan/vios:1.0

   docker container run -it --name vios --hostname vioser -v `pwd`/VIOS:/vios/VIOS -v `pwd`/database:/vios/database --link mysql:db suyanan/vios:1.0
   docker inspect -f '{{.NetworkSettings.IPAddress}}' vios
   #add IP manually, IF IP is not included int *ALLOWED_HOSTS* variable.
   ALLOWED_HOSTS = [
   u'127.0.0.1',
   u'localhost',
   u'172.17.0.3',
   ]
   

   FIANLLY, access IP:8000 in browser.

   CAUTION：

   link para is useful but not recommended in current Docker version(1.17).

1. generate network

   docker network create neter
   

2. start mysql container, and appoint the network

   docker run -it --hostname=mysqler --network neter --name mysql -e MYSQL_ROOT_PASSWORD=1234aaaa -e MYSQL_DATABASE=vios_db mysql:5.7
   

3. start vios container，and make its network same with that of mysql container(both network is neter)，achieve the access to MySQL create image, first：

   docker image build -t suyanan/vios:1.0 . docker pull suyanan/vios:1.0

   docker container run -it --name vios --hostname vioser --network neter -v `pwd`/VIOS:/vios/VIOS -v `pwd`/database:/vios/database suyanan/vios:1.0
   

   docker inspect mysql
   docker inspect vios
   

4. set the HOST elemen of DATABASE variable and ALLOWED_HOSTS variable in settings.py:

   DATABASES = {
       'default': {
           'ENGINE': 'django.db.backends.mysql',
           'NAME': 'vios_db',
           'HOST': ##mysql_container_ip
           'PORT':'3306',
           'USER':'root',
           'PASSWORD':'1234aaaa',
       }
   }
   ALLOWED_HOSTS = [
   u'127.0.0.1',
   u'localhost',
   #vios_container_ip
   ]
   

   CAUTION：

   • Just RESTART the vios container if you can't find sample_list or sample_se_list AFTER you upload your raw FASTQ datas.
   • start container with two commands:

     dockecontainer run -it --name vios --hostname vioser --network neter -v `pwd`/VIOS:/vios/VIOS -v `pwd`/database:/vios/database vios_pipeliner sh -c "python /vios/VIOS/manage.py migrate && python /vios/VIOS/manage.py runserver 0.0.0.0:8000"
     

   • start container with /bin/bash to test internally(cover the CMD in Dockerfile):

     dockecontainer run -it --name vios --hostname vioser --network neter -v `pwd`/VIOS:/vios/VIOS -v `pwd`/database:/vios/database vios_pipeliner /bin/bash
     

```
docker tag suyanan/vios：1.0 vioser_pipeliner
OR:
docker-compose up  #create and start containers
docker-compose down -v   #shutdown and remove containers
```

**CAUTION**：
  - There are still problems with using compose to manage the two containers(pipelinre——same as vios, and mysql)：
    *pipeliner_1  | django.db.utils.OperationalError: (2003, "Can't connect to MySQL server on 'mysqler' (111)")*

1. local images：

   docker image save suyanan/vios:x.0 -o suyanan_vios_x.0.tar
   docker image load -i suyanan_vios_x.tar
   

2. warning messages while building Dckerfile：

   WARNING: apt does not have a stable CLI interface. Use with caution in scripts. debconf: delaying package configuration, since apt-utils is not installe

3. NO supporting window interface WITH Docker deployment SO：Force matplotlib to not use any Xwindows backend. add codes below in file config_para.py：

   import matplotlib
   matplotlib.use('Agg')
   

   The visualization results is located in corresponding directories.

4. softwares downloaded postion：

   1.0/directory : JAVA and ngs softwares is necessary，and you can download from BaiduYun;

   2.0/directory : velvet is necessary，but you need to decompress and rename with "velvet_1.2.10.tgz", and recompress it.

201712.11

The End