limour-blog.github.io/source/_posts/2023-07-09-【记录】安装生信的代码编写环境.md

---

title: 【记录】安装生信的代码编写环境 urlname: -ji-lu--an-zhuang-sheng-xin-de-dai-ma-bian-xie-huan-jing date: 2023-07-09 20:01:33

tags: ['生信', 'docker', 'vscode', 'rstudio', 'jupyter']

安装 vscode-web

• 安装好 docker

• QUIC 转 SOCKS5

  mkdir -p ~/app/vscode && cd ~/app/vscode && nano docker-compose.yml
  sudo docker-compose up -d && sudo docker-compose logs
  

  version: "2.1"
  services:
  code-server:
  image: linuxserver/code-server:latest
  container_name: code-server
  environment:
    - PUID=1000
    - PGID=1000
    - TZ=Asia/Shanghai
    - PASSWORD=password
    - SUDO_PASSWORD=password
    - PROXY_DOMAIN=code-server.my.domain #optional
    - DEFAULT_WORKSPACE=/config/workspace #optional
  volumes:
    - ./config:/config
  ports:
    - 2441:8443
  restart: unless-stopped
  

  配置代理和中文

• 打开Visual Studio Code，点击Manage，在列表中选择Settings

• 在弹出的搜索框中输入"proxy"，即可看到代理的配置项"Http:Proxy"

• 宿主机获取docker0的ip: ip address | grep docker0

• 然后docker内设置代理 http://docker0的ip:port 拓展内搜索zh-cn，安装中文界面拓展

  安装 conda

• 回到WORKSPACE，ctrl+~ 调出终端

• sudo sed -i 's/archive.ubuntu.com/mirrors.aliyun.com/g' /etc/apt/sources.list

• sudo apt update

• sudo apt install wget

• 安装 conda

  安装nodejs

  conda create -n node -c conda-forge nodejs
  conda activate node
  npm config set registry https://registry.npmmirror.com
  

  使用git

  npm create astro@latest
  git config --global user.email "youremail"
  git config --global user.name "yourname"
  git branch -M main && git add . && git commit -m 'Initial commit'
  git remote add origin https://github.com/Limour-dev/chatGPT.git
  git push --set-upstream origin main --force # Creating a personal access token
  git config --global credential.helper cache
  git push
  

  hello world

  ---
  const search = Astro.url.searchParams.get('search')! || '';
  ---
  <h1>{search}</h1>
  

• Enabling SSR in Your Project

• 编辑 chatGPT/src/pages/index.astro

• npm run dev

• 访问 https://vscode.domain/proxy/3000/?search=hello%20world 进行测试

  安装 Jupyter

  持久化镜像存储

  mkdir -p ~/datascience && cd ~/datascience
  nano docker-compose.yml
  sudo docker-compose up -d
  sudo docker-compose logs
  sudo docker cp -a jupyterR:/opt /home/limour/upload/opt
  sudo docker cp -a jupyterR:/home/jovyan /home/limour/upload/home
  sudo docker-compose down && sudo docker volume prune
  

  version: '3.3'
  services:
  datascience-notebook:
      ports:
          - '57002:8888'
      container_name: jupyterR
      restart: always
      image: 'jupyter/datascience-notebook:r-4.3.1'
      command: start-notebook.sh --NotebookApp.token='***'
  

  启动镜像

  nano docker-compose.yml
  sudo chmod 777 -R /home/limour/upload/
  sudo docker-compose up -d
  sudo docker-compose logs
  

  version: '3.3'
  services:
  datascience-notebook:
      ports:
          - '57002:8888'
      container_name: jupyterR
      restart: always
      volumes:
          - '/home/limour/upload:/home/jovyan/upload'
          - '/home/limour/upload/opt:/opt'
          - '/home/limour/upload/home:/home/jovyan'
      image: 'jupyter/datascience-notebook:r-4.3.1'
      command: start-notebook.sh --NotebookApp.token='***'
  

• NPS内网穿透加NPM面板反代

  R包镜像

  nano .Rprofile
  options()$repos ## 查看使用install.packages安装时的默认镜像
  options()$BioC_mirror ##查看使用bioconductor的默认镜像
  

  options(BioC_mirror="https://mirrors.ustc.edu.cn/bioc/") ##指定镜像，这个是中国科技大学镜像
  options("repos" = c(CRAN="https://mirrors.tuna.tsinghua.edu.cn/CRAN/")) ##指定install.packages安装镜像，这个是清华镜像
  options(ggrepel.max.overlaps = Inf)
  

  安装 R 内核

  conda create -n seurat -c conda-forge r-seurat=4.1.1 -y
  conda activate seurat
  conda install -c conda-forge r-tidyverse -y
  conda install -c conda-forge r-irkernel -y
  Rscript -e "IRkernel::installspec(name='seurat', displayname='r-seurat')"
  conda install -c conda-forge r-devtools -y
  Rscript -e "BiocManager::install('glmGamPoi')"
  wget -e "https_proxy=http://172.17.0.1:8580" https://github.com/chris-mcginnis-ucsf/DoubletFinder/archive/refs/heads/master.zip -O DoubletFinder-master.zip
  Rscript -e "devtools::install_local('DoubletFinder-master.zip')"
  

  安装 python 内核

  conda create -n markdown2pptx -c conda-forge python -y
  conda install -n markdown2pptx ipykernel -c conda-forge -y
  conda run -n markdown2pptx python -m ipykernel install --user --name markdown2pptx
  conda run -n markdown2pptx pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple
  

  安装 Golang 内核

• 项目地址 gophernotes

  conda create -n golang -c conda-forge go -y
  conda activate golang
  go env -w GO111MODULE=on
  go env -w GOPROXY=https://mirrors.aliyun.com/goproxy/
  go install github.com/gopherdata/gophernotes@v0.7.5 #去仓库查看最新版本号
  mkdir -p ~/.local/share/jupyter/kernels/golang
  cd ~/.local/share/jupyter/kernels/golang
  cp "$(go env GOPATH)"/pkg/mod/github.com/gopherdata/gophernotes@v0.7.5/kernel/* "."
  chmod +w ./kernel.json
  sed "s|gophernotes|$(go env GOPATH)/bin/gophernotes|" < kernel.json.in > kernel.json
  

  安装 jupyter-ai 内核

• 获取端点和令牌

• 填写端点时尾部需要加上 /v1

  conda create -n jupyter-ai -c conda-forge jupyter-ai-magics -y
  conda install -n jupyter-ai -c conda-forge openai -y
  conda install -n jupyter-ai ipykernel -c conda-forge -y
  conda run -n jupyter-ai python -m ipykernel install --user --name jupyter-ai
  # pip install --upgrade jupyterlab -i https://pypi.tuna.tsinghua.edu.cn/simple # 升级到 JupyterLab 4
  # pip install jupyter_ai -i https://pypi.tuna.tsinghua.edu.cn/simple
  

• 启用拓展并配置端点和令牌

  %load_ext jupyter_ai_magics
  %env OPENAI_API_BASE=https://xxx/v1
  %env OPENAI_API_KEY=sk-xxx
  

• 示例

  %%ai chatgpt -f code
  解决汉诺塔问题的函数
  

  

安装 rstudio

mkdir -p ~/app/rstudio && cd ~/app/rstudio && nano docker-compose.yml
sudo docker-compose up -d && sudo docker-compose logs

version: '3'
services:
  rstudio:
    image: dceoy/rstudio-server
    container_name: Rstudio
    deploy:
      resources:
        limits:
          cpus: '0.50'
          memory: 500M
        reservations:
          cpus: '0.25'
          memory: 200M
    restart: always
    ports:
      - 57022:8787
    volumes:
      - /home/gene/zl_liu/rstudio:/home/rstudio
      - /home/gene/upload:/home/rstudio/upload
    working_dir: /home/rstudio

更改 R 版本

• 安装 conda

  # 容器内
  conda create -n r_4_1_3 -c conda-forge r-base=4.1.3 -y
  conda activate r_4_1_3
  whereis R
  # /home/rstudio/miniconda3/envs/r_4_1_3/bin/R
  # 容器外
  docker exec -it Rstudio /bin/bash
  chmod 777 -R  /etc/rstudio/
  exit 
  nano -K /etc/rstudio/rserver.conf
  # Server Configuration File
  rsession-which-r=/home/rstudio/miniconda3/envs/r_4_1_3/bin/R
  sudo docker-compose restart
  

  安装 seurat

  # 进入terminal，以下操作均在terminal中进行
  export R_LIBS_SITE=""
  # 在terminal中进入R
  .libPaths('/home/rstudio/miniconda3/envs/r_4_1_3/lib/R/library')
  .libPaths() 确保没有其他路径
  remove.packages("vctrs")
  install.packages("vctrs")
  install.packages('Seurat')
  remove.packages("cli")
  install.packages("cli")
  install.packages("tidyverse")
  install.packages("plotly")
  重启R session
  library(tidyverse)
  library(Seurat)
  

  绘制 3D-umap

  library(plotly)
  library(Seurat)
  sample13 <- readRDS("~/upload/zl_liu/work/Prognosis/scRNA/sample13.rds")
  sample13 <- RunUMAP(sample13, dims = 1:10, n.components = 3L)
  plot.data <- FetchData(object = sample13, vars = c("UMAP_1", "UMAP_2", "UMAP_3", "seurat_clusters"))
  plot.data$label <- paste(rownames(plot.data))
  # Plot your data, in this example my Seurat object had 21 clusters (0-20)
  plot_ly(data = plot.data, 
      x = ~UMAP_1, y = ~UMAP_2, z = ~UMAP_3, 
      color = ~seurat_clusters, 
      colors = c("lightseagreen",
                 "gray50",
                 "darkgreen",
                 "red4",
                 "red",
                 "turquoise4",
                 "black",
                 "yellow4",
                 "royalblue1",
                 "lightcyan3",
                 "peachpuff3",
                 "khaki3",
                 "gray20",
                 "orange2",
                 "royalblue4",
                 "yellow3",
                 "gray80",
                 "darkorchid1",
                 "lawngreen",
                 "plum2",
                 "darkmagenta")[1:7],
      type = "scatter3d", 
      mode = "markers", 
      marker = list(size = 5, width=2), # controls size of points
      text=~label, #This is that extra column we made earlier for which we will use for cell ID
      hoverinfo="text") #When you visualize your plotly object, hovering your mouse pointer over a point shows cell names
  

安装 notebook

mkdir -p ~/app/notebook && cd ~/app/notebook && nano docker-compose.yml
sudo docker-compose up -d && sudo docker-compose logs

version: '3'
services:
  notebook:
    image: jupyter/r-notebook
    container_name: notebook
    restart: unless-stopped
    environment:
      - DOCKER_STACKS_JUPYTER_CMD=nbclassic
      - TZ=Asia/Shanghai
      - JUPYTER_GATEWAY_REQUEST_TIMEOUT=600
      - JUPYTER_GATEWAY_CONNECT_TIMEOUT=600
    ports:
      - '57002:8888'
    volumes:
      - '/home/limour/upload:/home/jovyan/upload'
      - '/home/limour/upload/home:/home/jovyan'
    command: start-notebook.py --NotebookApp.token='***'

conda init
source activate
conda config --add envs_dirs /home/jovyan/upload/envs

proxy_connect_timeout   600s;
proxy_read_timeout      600s;
proxy_send_timeout      600s;
proxy_buffering         off;