eggNOG-mapper：功能注释集大成者

安装

Home · eggnogdb/eggnog-mapper Wiki · GitHub

git clone https://github.com/eggnogdb/eggnog-mapper.git
cd eggnog-mapper
python setup.py install
# 如果您的 eggnog-mapper 路径是/home/user/eggnog-mapper
export PATH=/home/user/eggnog-mapper:/home/user/eggnog-mapper/eggnogmapper/bin:"$PATH"
# 设置database安装路径
export EGGNOG_DATA_DIR=/home/user/eggnog-mapper-data
# 下载默认数据库
download_eggnog_data.py 
# 构建特定物种数据库
create_dbs.py -m diamond --dbname Viruses --taxids 2 #细菌的
#hmmer太慢了，舍弃
########################################################################################
# 下载数据库（我这里下载了hmmer的病毒数据库加上默认数据库）
# 更多可见 http://eggnog5.embl.de/#/app/downloads 和 查看download_eggnog_data.py --help
#download_eggnog_data.py -y -P -H -d 10239 --dbname Viruses
# 顺便把hmmer的其他几个关键数据库下载了
#nohup download_eggnog_data.py -y -H -d 2 --dbname Bacteria
#nohup download_eggnog_data.py -y -H -d 2759 --dbname Eukaryota
#nohup download_eggnog_data.py -y -H -d 2157 --dbname Archaea

使用 

python ~/Software/eggnog-mapper/emapper.py -i A4.2.1_wanzheng.faa -o 13 --cpu 90 --usemem --override --evalue 1e-5 --score 50 -m hmmer -d Viruses

• -i A4.2.1_wanzheng.faa：输入文件，应为一个包含蛋白质序列的FASTA文件。
• -o 13：输出文件的前缀，这里是13。EggNOG-mapper会生成多个输出文件，如13.emapper.annotations。
• --cpu 90：使用的CPU核心数，这里是90。
• --usemem：这个参数表示在搜索过程中使用尽可能多的内存。
• --override：如果输出文件已经存在，这个参数会让EggNOG-mapper覆盖它们。
• --evalue 1e-5：设置E值阈值，这里是1e-5。E值越小，表示匹配结果的可信度越高。
• --score 50：设置比对得分阈值，这里是50。得分越高，表示匹配结果的相似度越高

结果文件是emapper.annotations结尾的这个

参考别人的解释：基因组注释（6）——在线版eggNOG-mapper注释功能基因 - 我的小破站 

 我想写一个代码，对多个进行了eggnog注释的结果文件，根据结果的不同列，进行合并和计数

#! /usr/bin/env python
#########################################################
# make eggnog result together
# written by PeiZhong in IFR of CAAS

import argparse
import os
import pandas as pd

parser = argparse.ArgumentParser(description='TPM count file combine')
parser.add_argument('--input_dir', "-dir", required=True, help='Path to eggnog result files')
parser.add_argument('--marker', "-m", nargs='+', required=True, help='marker that can identify result files')
parser.add_argument('--out_prefix', required=True, help='output file prefix')
parser.add_argument('--level', '-l', type=str, required=True, help='COG_category, Description, GOs, EC, KEGG_ko, KEGG_Pathway, KEGG_Module, CAZy, PFAMs, only use one in a time')

args = parser.parse_args()

input_dir = args.input_dir
markers = args.marker
out_prefix = args.out_prefix
level = args.level

ls_level = ["COG_category","Description","GOs","EC","KEGG_ko","KEGG_Pathway","KEGG_Module","CAZy","PFAMs"]
if level not in ls_level:
    raise ValueError("Invalid level specified. Use COG_category, Description, GOs, EC, KEGG_ko, KEGG_Pathway, KEGG_Module, CAZy, PFAMs.")

result_ls = []

# Step 1: find result files
print(f"Scanning input directory: {input_dir}")
all_files = os.listdir(input_dir)
filtered_files = [
    os.path.join(input_dir, file)
    for file in all_files
    if all(marker in file for marker in markers)
]

if not filtered_files:
    print(f"No files matched the makers: {' '.join(markers)} in directory {input_dir}")
    exit(1)

# Step 2: Read and process files
result_df = pd.DataFrame()
temp_files = []  # List to track temporary files

for file in filtered_files:
    print(f"Preprocessing file: {file}")
    temp_file = f"{file}.temp"
    temp_files.append(temp_file)
    with open(file, "r") as infile, open(temp_file, "w") as outfile:
        for line in infile:
            if line.startswith("#query"):
                outfile.write(line.lstrip("#")) 
            else:
                outfile.write(line)

    sample_name = os.path.basename(file)  # Use the filename as the sample name
    print(f"Processing file: {temp_file}")
    df = pd.read_csv(temp_file, sep='\t', comment='#')  # Use the preprocessed temporary file
    if level not in df.columns:
        print(f"Level '{level}' not found in file {temp_file}. Skipping.")
        continue

    # Split comma-separated annotations into separate rows
    df = df[[level]].dropna()
    df[level] = df[level].str.split(',')  # Split annotations by commas
    df = df.explode(level)  # Expand the DataFrame so each annotation has its own row
    df[level] = df[level].str.strip("\n")  # Remove extra spaces if present

    # Count annotations for this sample
    count_df = df[level].value_counts().reset_index()
    count_df.columns = [level, sample_name]

    # Merge the count table into the result DataFrame
    if result_df.empty:
        result_df = count_df
    else:
        result_df = pd.merge(result_df, count_df, on=level, how='outer')

# Step 3: Fill missing values with 0 and output result
result_df.fillna(0, inplace=True)  # Replace NaN with 0 for counts
result_df = result_df.sort_values(by=level)  # Optional: Sort by annotation name

output_file = f"{out_prefix}_{level}_counts_matrix.csv"
print(f"Writing output to {output_file}")
result_df.to_csv(output_file, index=False)

# Step 4: Clean up temporary files
print("Cleaning up temporary files...")
for temp_file in temp_files:
    if os.path.exists(temp_file):
        os.remove(temp_file)

print("Done!")

eggnog_level_together.py -dir . -m emapper.annotations --out_prefix 16_ANF_genes -l CAZy

结果：