二代测序宏基因组分析软件安装及数据库下载

官方主页：http://huttenhower.sph.harvard.edu/
biobakery 主页：https://github.com/biobakery
Galaxy 主页：http://huttenhower.sph.harvard.edu/galaxy/

官方主页：http://segatalab.cibio.unitn.it/
github 主页：https://github.com/biobakery

开发维护软件：http://segatalab.cibio.unitn.it/tools/index.html
MetaPhlAn (2022 )，PhyloPhlAn (2020)，ViromeQC (2019)，curatedMD (2017)，MetaMLST 
(2016)，StrainPhlAn (2022)，MetAML (2016)，PanPhlAn (2020)，GraPhlAn (2015)，MetaRef 
(2014)，MetaPhlAn (2012)，LEfSe (2011)，ShortBRED (2015)，MicroPITA (2014)，HUMAnN 
(2012)

#bioBakery：http://segatalab.cibio.unitn.it/tools/biobakery/index.html
conda create -n biobakery python=3.7 -y
conda activate biobakery
conda config --add channels biobakery

#安装软件humann,kneaddata
conda install -y -c bioconda humann
#conda install -y -c biobakery humann 或者这个点安装
conda install -y -c biobakery kneaddata
conda install -y krona
conda install -y -c biobakery hclust2

#python2.7环境
conda create -n graphlan python=2.7 -y
conda activate graphlan
conda install -y graphlan
conda install -y export2graphlan
#conda install -y -c bioconda strainphlan  没安装上

$ kneaddata_database
KneadData Databases ( database : build = location )
human_genome : bowtie2 = http://huttenhower.sph.harvard.edu/kneadData_databases/Homo_sapiens_hg37_and_human_contamination_Bowtie2_v0.1.tar.gz
human_genome : bmtagger = http://huttenhower.sph.harvard.edu/kneadData_databases/Homo_sapiens_BMTagger_v0.1.tar.gz
human_transcriptome : bowtie2 = http://huttenhower.sph.harvard.edu/kneadData_databases/Homo_sapiens_hg38_transcriptome_Bowtie2_v0.1.tar.gz
ribosomal_RNA : bowtie2 = http://huttenhower.sph.harvard.edu/kneadData_databases/SILVA_128_LSUParc_SSUParc_ribosomal_RNA_v0.2.tar.gz
mouse_C57BL : bowtie2 = http://huttenhower.sph.harvard.edu/kneadData_databases/mouse_C57BL_6NJ_Bowtie2_v0.1.tar.gz
#可以将以上链接复制去本地迅雷下载，再上传到服务器 比较快
kneaddata_database --download human_genome bowtie2 ./ #自带命令下载比较慢
#第二种：国家微生物科学中心下载，挺快
lftp ftp://download.nmdc.cn/tools/kneaddata
mirror human_genome

#自建bowtie2索引
#下载拟南芥
axel -n 100 http://ftp.ensemblgenomes.org/pub/plants/release-53/fasta/arabidopsis_thaliana/dna/Arabidopsis_thaliana.TAIR10.dna.toplevel.fa.gz
gunzip Arabidopsis_thaliana.TAIR10.dna.toplevel.fa.gz
mv Arabidopsis_thaliana.TAIR10.dna.toplevel.fa tair10.fa
# bowtie2构建索引
bowtie2-build -f tair10.fa tair10 --threads 12

#安装 metaphlan
conda create --name mpa -c bioconda python=3.7 metaphlan
#官网下载，
Zenodo：https://zenodo.org/record/3957592#.YrFQd4dBwQ8
    wget https://zenodo.org/record/3957592/files/mpa_v30_CHOCOPhlAn_201901_marker_info.txt.bz2
    wget https://zenodo.org/record/3957592/files/mpa_v30_CHOCOPhlAn_201901.tar

Segatalab FTP：http://cmprod1.cibio.unitn.it/biobakery3/metaphlan_databases/
    http://cmprod1.cibio.unitn.it/biobakery3/metaphlan_databases/mpa_v31_CHOCOPhlAn_201901_marker_info.txt.bz2
    http://cmprod1.cibio.unitn.it/biobakery3/metaphlan_databases/mpa_v31_CHOCOPhlAn_201901.tar
    可以用迅雷下载完上传，两个文件共计400M左右。

#国家微生物科学中心下载，版本可能不对，建议还是上述迅雷下载
lftp ftp://download.nmdc.cn/tools/humann3
mirror metaphlan_databases/

#配置数据库
cd ~/Software/miniconda3/envs/mpa/lib/python3.7/site-packages/metaphlan/metaphlan_databases/
cp /share/home/xiehs/17.meta/database/metaphlan/* ./ #刚才下载文件所在的目录
    $ ll
    total 400M
    -rw-rw-r-- 2 xiehs xiehs   50 Oct 26 20:59 README.txt
    -rw-rw-r-- 1 xiehs xiehs  14M Oct 31 10:35 mpa_v31_CHOCOPhlAn_201901_marker_info.txt.bz2
    -rw-rw-r-- 1 xiehs xiehs 408M Oct 31 10:35 mpa_v31_CHOCOPhlAn_201901.tar

#方法1：从官网下载数据库，速度比较慢
#显示数据库内容
humann_databases
    http://huttenhower.sph.harvard.edu/humann_data/chocophlan/full_chocophlan.v296_201901b.tar.gz
#创建数据库目录
mkdir humann3_databases 
#方法1：官网下载，下载速度慢，用迅雷吧
humann_databases --download chocophlan full humann3_databases
humann_databases --download uniref uniref90_diamond humann3_databases
#humann_databases --download uniref uniref90_ec_filtered_diamond humann3_databases
humann_databases --download uniref uniref50_diamond humann3_databases
#humann_databases --download uniref uniref50_ec_filtered_diamond humann3_databases
humann_databases --download utility_mapping full humann3_databases

#方法2：国家微生物科学数据中心下载数据库，版本不对，这种方式下载不建议
lftp ftp://download.nmdc.cn/tools/
ls
mirror humann3 #full_chocophlan.v201901.tar.gz可见版本不对，不是前面humann_databases中的01b版本，这种方式下载不建议
mkdir chocophlan uniref utility_mapping

将4个数据库分别下载上传后，解压到各自目录

#3 更新数据库
humann_config --print
#数据库目录 /ifs1/MetaDatabase/humann3_databases，这里选用90的ref，准一些
humann_config --update database_folders nucleotide /share/home/xiehs/17.meta/database/humann3_databases/full_chocophlan
humann_config --update database_folders protein /share/home/xiehs/17.meta/database/humann3_databases/uniref90_diamond
humann_config --update database_folders utility_mapping /share/home/xiehs/17.meta/database/humann3_databases/utility_mapping
#更新完检查
humann_config --print
#修改线程
humann_config --update run_modes threads 12

bioinfoer.com