使用biopython解析kegg数据库

欢迎关注”生信修炼手册”!

KEGG数据库称之为基因组百科全书，是一个包含gene, pathway等多个子数据库的综合性数据库。为了更好的查询kegg数据，官方提供了对应的API。

在biopython中，通过Bio.KEGG模块，对kegg官方的API进行了封装，允许在python环境中使用kegg API。KEGG API与python代码的对应关系如下

/list/hsa:10458+ece:Z5100 -> REST.kegg_list(["hsa:10458", "ece:Z5100"])
/find/compound/300-310/mol_weight -> REST.kegg_find("compound", "300-310", "mol_weight")
/get/hsa:10458+ece:Z5100/aaseq -> REST.kegg_get(["hsa:10458", "ece:Z5100"], "aaseq")

利用REST模块，可以下载API支持的任何类型的数据，以pathway为例，示例如下

>>> from Bio.KEGG import REST
>>> pathway = REST.kegg_get('hsa00010')

对于查询获得的内容，通过read方法可以转换为纯文本，示例如下

>>> pathway = REST.kegg_get('hsa00010')
>>> res = pathway.read().split("\n")
>>> res[0]
'ENTRY hsa00010 Pathway'
>>> res[1]
'NAME Glycolysis / Gluconeogenesis - Homo sapiens (human)'
>>> res[2]



    
'DESCRIPTION Glycolysis is the process of converting glucose into pyruvate and generating small amounts of ATP (energy) and NADH (reducing power). It is a central pathway that produces important precursor metabolites: six-carbon compounds of glucose-6P and fructose-6P and three-carbon compounds of glycerone-P, glyceraldehyde-3P, glycerate-3P, phosphoenolpyruvate, and pyruvate [MD:M00001]. Acetyl-CoA, another important precursor metabolite, is produced by oxidative decarboxylation of pyruvate [MD:M00307]. When the enzyme genes of this pathway are examined in completely sequenced genomes, the reaction steps of three-carbon compounds from glycerone-P to pyruvate form a conserved core module [MD:M00002], which is found in almost all organisms and which sometimes contains operon structures in bacterial genomes. Gluconeogenesis is a synthesis pathway of glucose from noncarbohydrate precursors. It is essentially a reversal of glycolysis with minor variations of alternative paths [MD:M00003].'

这样就可以通过字符串解析，来获取通路对应的编号，名称，注释等信息。对于KEGG数据的解析，biopython还提供了专门的解析函数，但是解析函数并不完整，目前只覆盖了compound, map, enzyme等子数据库。以enzyme数据库为例，用法如下

>>> from Bio.KEGG import REST
>>> request = REST.kegg_get("ec:5.4.2.2")
>>> open("ec_5.4.2.2.txt", "w").write(request.read())
>>> records = Enzyme.parse(open("ec_5.4.2.2.txt"))
>>> record = list(records)[0]
>>> record
0x02EE7D18>
>>> record.classname
['Isomerases;', 'Intramolecular transferases;', 'Phosphotransferases (phosphomutases)']
>>> record.entry
'5.4.2.2'

通过biopython，我们不仅可以在python环境中使用kegg api, 更重要的是，可以借助python的逻辑处理，来实现复杂的筛选逻辑，比如查找human中DNA修复相关的基因，基本思路如下

1. 通过list API获取human所有的pathway编号；

2. 通过get API获取每条pathway, 解析其description信息，筛选出现了repair关键词的通路；

3. 对于筛选出的通路，通过文本解析获取该通路对应的基因；

完整的代码如下

>>> from Bio.KEGG import REST
>>> human_pathways = REST.kegg_list("pathway", "hsa").read()
>>> repair_pathways = []
>>> for line in human_pathways.rstrip().split("\n"):
...     entry, description = line.split("\t")
...     if "repair" in description:
...         repair_pathways.append(entry)
...
>>> repair_pathways



    
['path:hsa03410', 'path:hsa03420', 'path:hsa03430']
>>> repair_genes = []
>>> for pathway in repair_pathways:
...     pathway_file = REST.kegg_get(pathway).read()
...     current_section = None
...     for line in pathway_file.rstrip().split("\n"):
...         section = line[:12].strip()
...         if not section == "":
...             current_section = section
...         if current_section == "GENE":
...             gene_identifiers, gene_description = line[12:].split("; ")
...             gene_id, gene_symbol = gene_identifiers.split()
...             if not gene_symbol in repair_genes:
...                 repair_genes.append(gene_symbol)
...
>>> repair_genes
['OGG1', 'NTHL1', 'NEIL1', 'NEIL2', 'NEIL3', 'UNG', 'SMUG1', 'MUTYH', 'MPG', 'MBD4', 'TDG', 'APEX1', 'APEX2'


    
, 'POLB', 'POLL', 'HMGB1', 'XRCC1', 'PCNA', 'POLD1', 'POLD2', 'POLD3', 'POLD4', 'POLE', 'POLE2', 'POLE3', 'POLE4', 'LIG1', 'LIG3', 'PARP1', 'PARP2', 'PARP3', 'PARP4', 'FEN1', 'RBX1', 'CUL4B', 'CUL4A', 'DDB1', 'DDB2', 'XPC', 'RAD23B', 'RAD23A', 'CETN2', 'ERCC8', 'ERCC6', 'CDK7', 'MNAT1', 'CCNH', 'ERCC3', 'ERCC2', 'GTF2H5', 'GTF2H1', 'GTF2H2', 'GTF2H2C_2', 'GTF2H2C', 'GTF2H3', 'GTF2H4', 'ERCC5', 'BIVM-ERCC5', 'XPA', 'RPA1', 'RPA2', 'RPA3', 'RPA4', 'ERCC4'


    
, 'ERCC1', 'RFC1', 'RFC4', 'RFC2', 'RFC5', 'RFC3', 'SSBP1', 'PMS2', 'MLH1', 'MSH6', 'MSH2', 'MSH3', 'MLH3', 'EXO1']

通过biopython, 可以更加高效的使用KEGG API, 结合API的数据获取能力和python的逻辑处理能力，来满足我们的个性化分析需求。

·end·

—如果喜欢，快分享给你的朋友们吧—

原创不易，欢迎收藏，点赞，转发！生信知识浩瀚如海，在生信学习的道路上，让我们一起并肩作战！

本公众号深耕耘生信领域多年，具有丰富的数据分析经验，致力于提供真正有价值的数据分析服务，擅长个性化分析，欢迎有需要的老师和同学前来咨询。