噬菌体宿主预测工具——iPHoP

# Bioconda安装（推荐方式）
# 创建专属环境避免依赖冲突
conda create -n iPHoP -c conda-forge -c bioconda iphop
# 激活环境
conda activaste iPHoP
# 验证安装是否成功
iphop --version
# Docker容器部署
# 获取官方镜像
docker pull quay.io/biocontainers/iphop:v1.2.0
# 验证安装是否成功
docker run -it quay.io/biocontainers/iphop:v1.2.0 iphop --help

# 软件推荐的下载命令
iphop download --db_dir iphop_db
# 这个数据库最新的版本有279G，根据小编的经验，使用iphop下载速度慢，而且容易中断，官方已经将数据集拆分成27个子集，类似于GTDB数据库，小伙伴们可以先用wget批量下载，经过md5校验后cat命令进行合并
for i in {00..27}
do 
    wget -c -nv -b -a download.log https://portal.nersc.gov/cfs/m342/iphop/db/iPHoP_db_Jun25_rw.tar.gz.${i}
done
cat iPHoP_db_Jun25_rw.tar.gz* > iPHoP_db_Jun25_rw.tar.gz
# 解压
mkdir -p iPHoP_db_Jun25_rw
tar -xvzf iPHoP_db_Jun25_rw.tar.gz -C iPHoP_db_Jun25_rw
# 检验数据库
cd iPHoP_db_Jun25_rw
md5sum -c md5checkfile.txt > md5checkfile.check

# 查看软件使用帮助
iphop -h 
iPHoP v1.3.3: integrating Host Phage Predictions
https://bitbucket.org/srouxjgi/iphop
usage: iphop <task> [options]
task:
        predict         run full pipeline to generate host prediction for some input phage genome(s)
        download        download and setup the latest host database
        add_to_db       add some host genomes to the database of hosts (e.g. MAGs derived from the same metagenome(s))
other:
        split           small utility to split a fasta file into smaller batches (can be useful if you would like to split your input and run iPHoP separately for each batch)
        clean           small utility to clean the output directory of iPHoP by compressing some of the larger files.
optional arguments:
  -h, --help  show this help message and exit

# 病毒宿主预测
iphop predict -h 
Run full pipeline to predict host of input bacteriophage/archaeal virus genomes
usage: iphop predict --fa_file <input genomes> --out_dir <output directory> [options]
optional arguments:
  -h, --help            show this help message and exit
  --fa_file INPUT_FILE, -f INPUT_FILE
                        input fasta file of virus sequences
  --out_dir OUT_DIR, -o OUT_DIR
                        output directory (default = ./)
  --db_dir DB_PATH, -d DB_PATH
                        database directory (default = /home/Data/Jun25_2025_pub_rw/db)
  --num_threads THREADS, -t THREADS
                        number of threads (default = 1)
  --step STEP, -s STEP  To run only specific step(s) (default: all, options: compute / classify)
  --min_score MIN_SCORE, -m MIN_SCORE
                        Minimum confidence score for final output file - between 75 and 100. Default is 90, i.e. ~10% FDR (roughly). For ~5% overall (roughly) estimated FDR, use a minimum score of 95, etc.').
  --no_qc               Bypass the automated QC that filters out input sequences with > 10% Ns or with characters other than ATCGN. Warning: if set to true, low-quality sequences (with Ns or unusual characters) may lead to unexpected failures at multiple steps. (default = False)
  --debug               Run in a more verbose mode for debugging / troubleshooting purposes (warning: iPHoP becomes quite chatty in this mode..)
  --single_thread_wish  Force the recoded version of wish to run on a single thread instead of all the threads given to iPHoP (with --num_threads)

# 自定义宿主库
iphop add_to_db -h 
Add new genomes to the host database, for instance, new MAGs assembled from the same sample as the input phage/archaeovirus genomes
ote: the original database will not be duplicated but only linked to the updated one. If the path to the original database changes, the updated database will need to be rebuilt.
usage: iphop add_to_db --input_dir <directory with new host genomes> --gtdb_dir <output directory from GTDBtk (bacteria and archaea)> [options]
optional arguments:
  -h, --help            show this help message and exit
  --fna_dir FNA_DIR, -f FNA_DIR
                        input fasta directory (one fasta file per genome)
  --gtdb_dir GTDB_DIR, -g GTDB_DIR
                        GTDBtk result directory, with bacteria and/or archaea tree(s) including the new genomes to be added to the database
  --out_dir OUT_DIR, -o OUT_DIR
                        output directory - will include the new database (default = ./new_db)
  --num_threads THREADS, -t THREADS
                        number of threads (default = 1)
  --db_dir DB_PATH, -d DB_PATH
                        absolute path to the original database directory - will be completed with new genomes (default = /mnt/sdb/share/database/Tools/iPHoP/Jun25_2025_pub_rw/db)
  --debug               Run in a more verbose mode for debugging / troubleshooting purposes (warning: iPHoP becomes quite chatty in this mode..)

# 使用默认的数据库预测宿主
iphop predict --fa_file virus_assembly_genome.fna --out_dir iphop --db_dir iPHoP/Jun25_2025_pub_rw --num_threads 32 --min_score 90
# 添加binning好的MAGs当宿主(官方给的示例数据)
# GTDB构建系统发育信息(以细菌为例，这步很慢)
gtdbtk de_novo_wf --genome_dir Data_test_add_to_db_v1.4/Wetland_MAGs/ --bacteria --outgroup_taxon p__Patescibacteriota --out_dir Wetland_MAGs_GTDB-tk_results/ --cpus 32 --force --extension fa
# 将有注释信息和系统发育的信息添加到宿主数据库
iphop add_to_db --fna_dir Data_test_add_to_db_v1.4/Wetland_MAGs/ --gtdb_dir Wetland_MAGs_GTDB-tk_results/ --out_dir Jun_2025_pub_rw_w_Wetland_hosts --db_dir iPHoP/Jun_2025_pub_rw
# 使用新构建的数据库做预测
iphop predict --fa_file Input_viral_contigs.fasta --db_dir Jun_2025_pub_rw_w_Wetland_hosts/ --out_dir test_add_db -t 4 --min_score 75