融合跨物种科学数据的性状调控基因本体模型构建及应用

doi:10.13560/j.cnki.biotech.bull.1985.2023-0748

摘要/Abstract

摘要：

【目的】在新技术带来的育种数据激增与计算育种对知识服务的新需求下，为解决作物育种知识服务中跨物种学科知识获取效率低且优异多效基因发现困难的问题。【方法】本研究构建了性状调控基因本体模型框架，并定义了本体模型中的实体层次结构和实体属性。以主粮作物水稻、玉米、小麦和模式植物拟南芥为数据采集对象，构建了以性状调控基因本体模型为模式层的知识图谱并进行实验。【结果】最终形成了涵盖13种实体、16种数据属性和14个对象属性的性状调控基因本体模型，以此模型为本体层的知识图谱实现了跨物种间学科知识关联检索、优异多效基因挖掘和跨物种基因功能预测。【结论】本研究所提出的性状调控基因本体模型构建方法，能够实现跨物种间性状调控基因的关联发现，可提高跨物种学科知识的获取效率，可支撑多维度科学数据寻证分析的功能基因发现结果。本研究为多效基因的挖掘和基因功能预测提供了一条可实现的方法路径，为作物育种科学研究提供了有效的数据支撑服务。

关键词: 本体模型, 学科知识发现, 知识图谱, 跨物种, 调控基因

Abstract:

【Objective】 With the proliferation of breeding data brought by new technologies and the new demand for knowledge services in computational breeding, in order to solve the problem of inefficient cross-species subject knowledge acquisition and difficult discovery of elite pleiotropic genes in crop breeding knowledge service.【Method】 We constructed trait-regulated-genes ontology model framework, and defined entity hierarchical structure and entity attributes in ontology model. Taking staple crops rice, corn, wheat and model plant Arabidopsis thalliana as data collection objects, a knowledge graph with trait-regulated-genes ontology model as model layer is constructed and experimented.【Result】 Finally, the trait-regulated-genes ontology model covering 13 entities, 16 data attributes and 14 object attributes was formed. This model is used as the knowledge graph of ontology layer to realize cross-species subject knowledge association retrieval, mining of elite pleiotropic genes and prediction of gene function across species.【Conclusion】 The method of trait-regulated-genes ontology model construction proposed in our study may achieve the correlation discovery of trait regulatory genes across species, improve the efficiency of cross-species subject knowledge acquisition, and support the gene discovery results of multi-dimensional data analysis. This study provides a feasible method path for the mining of pleiotropic genes and gene function prediction, and provides effective data support services for crop breeding scientific research.

Key words: ontology model, knowledge discovery, knowledge graph, cross-species, regulatory gene

张丹丹, 赵瑞雪, 鲜国建, 熊赫. 融合跨物种科学数据的性状调控基因本体模型构建及应用[J]. 生物技术通报, 2024, 40(2): 313-324.

ZHANG Dan-dan, ZHAO Rui-xue, XIAN Guo-jian, XIONG He. Trait-regulated-genes Ontology Model Construction and Application by Integrating Cross-species Scientific Data[J]. Biotechnology Bulletin, 2024, 40(2): 313-324.

图/表 8

图1 性状调控基因本体模型框架

Fig. 1 Trait-regulated-genes ontology model framework

表1 性状调控基因本体模型中实体的释义

Table 1 Description of entity classes in trait-regulated-genes ontology model

数据类型维度 Data type dimension	实体名称 Entity name	数据属性 Data attribute
基因水平 Gene level	基因 Gene	基因标识符；物种；物理位置；转录本名称；PANTHER注释编号 Gene identity; species; location; transcript name; PANTHER identity
基因水平 Gene level	基因符号 Gene symbol	功能描述 Function description
蛋白水平 Protein level	蛋白 Protein	蛋白标识符；物种；首次被发现时间；功能描述；影响表型描述；文献编号 Protein identity; species; date of creation; function description; phenotype disruption; PubMed identity
	蛋白家族 Protein family	蛋白家族数据库编号；名称 Pfam identity; name
	酶 Enzyme	EC编号；名称 EC number; name
	结构域 Domain	名称；类型 Name; type
	亚细胞定位 Subcellular localization	名称；类型 Name; type
富集通路水平 Enrichment pathways level	生物学过程 Biological process	GO编号；名称 GO identity; name
	细胞组分 Cellular component	GO编号；名称 GO identity; name
	分子功能 Molecular function	GO编号；名称 GO identity; name
	代谢通路 Metabolic pathway	KO编号；名称 KO identity; name
	信号通路 Signal pathway	名称；类型 Name; type
性状水平 Trait level	性状 Trait	名称；类型 Name; type

表2 性状调控基因本体模型中实体数据属性释义

Table 2 Description of data attributes in trait-regulated-genes ontology model

数据属性名称 Data attributes	释义 Description
基因标识符 Gene identity	基因的唯一标识符用于区分基因类实例
转录本名称 Transcript name	描述基因在转录后所对应的名称
物理位置 Location	描述基因在染色体中的具体物理位点
PANTHER数据库编号 PANTHER identity	描述基因在PANTHER数据库中的具体注释信息编号
蛋白标识符 Protein identity	蛋白的唯一标识符用于区分蛋白类实例
首次被发现时间 Date of creation	描述蛋白初次被发现的时间
影响表型描述 Phenotype disruption	描述调控表型的主要分子机制
物种 Species	描述基因或蛋白所属于的物种
文献编号 PubMed identity	描述报道该蛋白的文献在PubMed数据库中的编号
功能描述 Function description	描述蛋白类实例主要的功能
蛋白家族数据库编号 Pfam identity	描述蛋白家族在Pfam数据库中的具体编号信息
EC编号 EC number	描述基因所编码酶的类型信息所属的EC编号
KO数据库编号 KO identity	描述基因在KO数据库中的具体注释信息编号
GO数据库编号 GO identity	描述基因的分子调控、细胞组分、生物学过程信息在GO数据库中所对应的编号
名称 Name	描述性状、蛋白家族、代谢途径等核心类的具体实例名称
类型 Type	描述性状、结构域、通路等核心类所属的类别名称

表3 性状调控基因本体模型中对象属性释义和数据来源

Table 3 Description of object attributes in trait-regulated-genes ontology model and data sources

对象属性名称Object attributes	描述对象 Object	数据来源 Data source	释义 Description
与……有关 Associates with	蛋白；性状Protein; trait	UniProt, PubMed	描述蛋白和性状之间的关联关系
与……同源 Homologous to	蛋白；蛋白 Protein ; protein	UniProt	描述蛋白与蛋白之间的同源关系
与......互作 Interacts with	蛋白；蛋白 Protein ; protein	UniProt, STRING	描述蛋白与蛋白间相互作用的关系
与……相对应 Corresponding to	蛋白；基因Protein; gene	UniProt, Phytozome, Ensembl plants, RGAP	描述基因和蛋白间的对应关系
与……一致 Identify with	蛋白；基因符号Protein; gene symbol	UniProt	描述蛋白和基因符号的关联关系
参与...... Involves in	蛋白；信号通路 Protein; signal pathway	UniProt	描述蛋白和信号通路的关联关系
表达于 Located in	蛋白；亚细胞定位位置 Protein; subcellular localization	UniProt	描述蛋白亚细胞定位的位置
具有......蛋白结构域 Has protein domain	蛋白；结构域Protein; domain	UniProt	描述蛋白所具有的蛋白结构域
属于...... Belongs to	蛋白；蛋白家族Protein; protein family	UniProt, Pfam	描述蛋白所属的蛋白家族
表达于 Located in	基因；细胞组分 Gene; cellular component	Phytozome, Ensembl plants, RGAP, KEGG	描述基因的亚细胞定位
行使功能 Performs	基因；分子功能 Gene; molecular function	Phytozome, Ensembl plants, RGAP, GO	描述基因在分子水平上所行使的功能
参与...... Involves in	基因；生物学过程 Gene; biological process	Phytozome, Ensembl plants, RGAP, GO	描述基因功能所直接参与的生物过程
参与..... Involves in	基因；代谢通路 Gene; metabolic pathway	Phytozome, Ensembl plants, RGAP, KEGG	描述基因在代谢通路中的功能
编码......类型的酶 Encodes the enzyme type	基因；酶Gene; enzyme	Phytozome, Ensembl plants, RGAP, KEGG	描述基因所编码蛋白所属的酶的类型

表3 性状调控基因本体模型中对象属性释义和数据来源

Table 3 Description of object attributes in trait-regulated-genes ontology model and data sources

对象属性名称Object attributes	描述对象 Object	数据来源 Data source	释义 Description
与……有关 Associates with	蛋白；性状Protein; trait	UniProt, PubMed	描述蛋白和性状之间的关联关系
与……同源 Homologous to	蛋白；蛋白 Protein ; protein	UniProt	描述蛋白与蛋白之间的同源关系
与......互作 Interacts with	蛋白；蛋白 Protein ; protein	UniProt, STRING	描述蛋白与蛋白间相互作用的关系
与……相对应 Corresponding to	蛋白；基因Protein; gene	UniProt, Phytozome, Ensembl plants, RGAP	描述基因和蛋白间的对应关系
与……一致 Identify with	蛋白；基因符号Protein; gene symbol	UniProt	描述蛋白和基因符号的关联关系
参与...... Involves in	蛋白；信号通路 Protein; signal pathway	UniProt	描述蛋白和信号通路的关联关系
表达于 Located in	蛋白；亚细胞定位位置 Protein; subcellular localization	UniProt	描述蛋白亚细胞定位的位置
具有......蛋白结构域 Has protein domain	蛋白；结构域Protein; domain	UniProt	描述蛋白所具有的蛋白结构域
属于...... Belongs to	蛋白；蛋白家族Protein; protein family	UniProt, Pfam	描述蛋白所属的蛋白家族
表达于 Located in	基因；细胞组分 Gene; cellular component	Phytozome, Ensembl plants, RGAP, KEGG	描述基因的亚细胞定位
行使功能 Performs	基因；分子功能 Gene; molecular function	Phytozome, Ensembl plants, RGAP, GO	描述基因在分子水平上所行使的功能
参与...... Involves in	基因；生物学过程 Gene; biological process	Phytozome, Ensembl plants, RGAP, GO	描述基因功能所直接参与的生物过程
参与..... Involves in	基因；代谢通路 Gene; metabolic pathway	Phytozome, Ensembl plants, RGAP, KEGG	描述基因在代谢通路中的功能
编码......类型的酶 Encodes the enzyme type	基因；酶Gene; enzyme	Phytozome, Ensembl plants, RGAP, KEGG	描述基因所编码蛋白所属的酶的类型

图2 性状调控基因知识图谱的层级知识结构 A：基因LOC_Os01g40094的多维度科学数据关联检索；B：蛋白P49597数据属性展示。图A中不同颜色的圆圈和箭头线分别代表不同的实体类型和关系类型，括号中的数字表示不同类型的节点和关系数量。以节点类型举例，红色圆圈代表“性状（trait）”节点，数量为3；浅蓝色圆圈代表“蛋白（protein）”节点，数量为4；橙色圆圈代表“基因（gene）”节点，数量为2。以关系类型举例，浅绿色箭头线代表“具有的蛋白结构域（has protein domain）”关系类型，数量为1；深绿色箭头线代表“属于（belongs to）”关系类型，数量为2

Fig. 2 Hierarchical knowledge structure of trait-regulated-genes knowledge graph A: Multi-dimensional scientific data association retrieval of gene LOC_Os01g40094. B: Data attributes of protein P49597. Different colored circles and arrow lines represent different entity types and relationship types in figure A, respectively. The numbers in parentheses indicate the number of nodes and relationships of different types. Taking node types as an example, red circles represent “trait” nodes, and the number is 3. The light blue circle represents the “protein” node, the number is 4. The orange circle represents the “gene” node, and the number is 2. For example, taking relationship types as an example, the light green arrow line represents the “has protein domain” relation type, and the number is 1. The dark green arrow line represents the “ belongs to “ relation type with a number of 2

图3 基因调控性状的预测图中不同颜色的圆圈和箭头线分别代表不同的实体类型和关系类型，括号中的数字表示不同类型的节点和关系数量。以节点类型举例，红色圆圈代表“性状（trait）”节点，数量为1；浅蓝色圆圈代表“蛋白（protein）”节点，数量为7；橙色圆圈代表“基因（gene）”节点，数量为7。以关系类型举例，浅绿色箭头线代表“具有的蛋白结构域（has protein domain）”关系类型，数量为6；深绿色箭头线代表“属于（belongs to）”关系类型，数量为7。本图中显示基因“TraesCS4B02G060000”节点和性状“plant height”节点间的关联路径共有8条

Fig. 3 Prediction of gene-regulated- trait The circles and arrow lines in different colors in the figure represent different entity types and relationship types respectively, and the numbers in brackets represent the number of nodes and relationships of different types. Taking node types as an example, the red circle represents a “trait” node and the number is 1. The light blue circle represents the “protein” node, the number of which is 7. The orange circle represents the “gene” node, with a number of 7. Take relationship types as an example, the light green arrow line represents the “has protein domain” relationship type, the number is 6. The dark green arrow line represents the “ belongs to” relation type with a number of 7. This figure shows that there are 8 association paths between the gene “TraesCS4B02G060000” node and the trait “plant height” node

图4 多效基因TraesCS3D02G078500的挖掘图中不同颜色的圆圈和箭头线分别代表不同的实体类型和关系类型，括号中的数字表示不同类型的节点和关系数量。以节点类型举例，红色圆圈代表“性状（trait）”节点，数量为5；浅蓝色圆圈代表“蛋白（protein）”节点，数量为5；橙色圆圈代表“基因（gene）”节点），数量为5。以关系类型举例，橘色箭头线代表“定位于（located in）”关系类型，数量为5；橙色箭头线代表“相对应（corresponding to）”关系类型，数量为5。本图预测基因“TraesCS3D02G078500”可能调控抗病“drought resistance”、抗盐“salt resistance”、粒重“grain weight”、株高“plant height”性状

Fig. 4 Mining of the elite pleiotropy gene TraesCS3D02G078500 The circles and arrow lines in different colors in the figure represent different entity types and relationship types respectively, and the numbers in brackets represent the number of nodes and relationships of different types. Taking node types as an example, the red circle represents the trait node and the number is 5. The light blue circle represents the “protein” node, the number of which is 5. The orange circle represents the “gene” node, and the number is 5. Take relationship types as an example, the orange arrow line represents the “located in” relationship type, and the number is 5. The yellow red arrow line represents the “ corresponding to “ relationship type, and the number is 5. This figure predicts that the gene “TraesCS3D02G078500” may regulate the traits of drought resistance, salt resistance, grain weight and plant height

图5 跨物种基因功能预测图中不同颜色的圆圈和箭头线分别代表不同的实体类型和关系类型，括号中的数字表示不同类型的节点和关系数量。以节点类型举例，红色圆圈代表“性状（trait）”节点，数量为1；浅蓝色圆圈代表“蛋白（protein）”节点，数量为10；橙色圆圈代表“基因（gene）”节点，数量为4。以关系类型举例，橘色箭头线代表“定位于（located_in）”关系类型，数量为13；橙色箭头线代表“相对应（corresponding_to）”关系类型，数量为10。本图建立了小麦基因“TraesCS2D02G261300”与拟南芥基因“AT1G48520”、水稻基因“LOC_Os11g34210”、玉米基因“Zm00001d052622”节点之间的关联路径

Fig. 5 Prediction of gene function across species The circles and arrow lines in different colors in the figure represent different entity types and relationship types respectively, and the numbers in brackets represent the number of nodes and relationships of different types. Taking node types as an example, the red circle represents a “trait” node and the number is 1. The light blue circle represents the “protein” node, with a number of 10. The orange circle represents the “gene” node, with a number of 4. Take relationship types as an example, the orange arrow line represents the “located in” relationship type, and the number is 13. The yellow red arrow line represents the “ corresponding to “ relationship type, and the number is 10. This figure establishes the association path between wheat gene “TraesCS2D02G261300” and Arabidopsis gene “AT1G48520”, rice gene “LOC_Os11g34210” and maize gene “Zm00001d052622”

参考文献 29

[1]	Wallace JG, Rodgers-Melnick E, Buckler ES. On the road to breeding 4.0: unraveling the good, the bad, and the boring of crop quantitative genomics[J]. Annu Rev Genet, 2018, 52: 421-444. doi: 10.1146/annurev-genet-120116-024846 pmid: 30285496
[2]	Goble C, Stevens R. State of the nation in data integration for bioinformatics[J]. J Biomed Inform, 2008, 41(5): 687-693. doi: 10.1016/j.jbi.2008.01.008 pmid: 18358788
[3]	Hassani-Pak K, Castellote M, Esch M, et al. Developing integrated crop knowledge networks to advance candidate gene discovery[J]. Appl Transl Genom, 2016, 11: 18-26. doi: 10.1016/j.atg.2016.10.003 pmid: 28018846
[4]	Ashburner M, Ball CA, Blake JA, et al. Gene ontology: tool for the unification of biology[J]. Nat Genet, 2000, 25(1): 25-29. doi: 10.1038/75556 pmid: 10802651
[5]	Zeeberg BR, Feng WM, Wang G, et al. GoMiner: a resource for biological interpretation of genomic and proteomic data[J]. Genome Biol, 2003, 4(4): R28. doi: 10.1186/gb-2003-4-4-r28 pmid: 12702209
[6]	Antezana E, Egaña M, Blondé W, et al. The Cell Cycle Ontology: an application ontology for the representation and integrated analysis of the cell cycle process[J]. Genome Biol, 2009, 10(5): R58. doi: 10.1186/gb-2009-10-5-r58 URL
[7]	Shrestha R, Arnaud E, Mauleon R, et al. Multifunctional crop trait ontology for breeders'data: field book, annotation, data discovery and semantic enrichment of the literature[J]. AoB Plants, 2010, 2010: plq008.
[8]	Shrestha R, Matteis L, Skofic M, et al. Bridging the phenotypic and genetic data useful for integrated breeding through a data annotation using the Crop Ontology developed by the crop communities of practice[J]. Front Physiol, 2012, 3: 326. doi: 10.3389/fphys.2012.00326 pmid: 22934074
[9]	Cooper L, Walls RL, Elser J, et al. The plant ontology as a tool for comparative plant anatomy and genomic analyses[J]. Plant Cell Physiol, 2013, 54(2): e1. doi: 10.1093/pcp/pcs163 URL
[10]	Hassani-Pak K, Singh A, Brandizi M, et al. KnetMiner: a comprehensive approach for supporting evidence-based gene discovery and complex trait analysis across species[J]. Plant Biotechnol J, 2021, 19(8): 1670-1678. doi: 10.1111/pbi.13583 pmid: 33750020
[11]	Venkatesan A, Tagny Ngompe G, Hassouni NE, et al. Agronomic Linked Data(AgroLD): a knowledge-based system to enable integrative biology in agronomy[J]. PLoS One, 2018, 13(11): e0198270. doi: 10.1371/journal.pone.0198270 URL
[12]	Momtchev V, Peychev D, Primov T, et al. Expanding the pathway and interaction knowledge in linked life data[J]. Int Semant Web Chall, 2009. http://challenge.semanticweb.org/.
[13]	隗玲, 胡正银, 庞弘燊, 等. 基于“主语-谓语-宾语”三元组的知识发现研究——以诱导多能干细胞领域为例[J]. 数字图书馆论坛, 2017(9): 28-34.
	Wei L, Hu ZY, Pang HS, et al. Study on knowledge discovery based on “subject-predication-object” predications: a case study of induced pluripotent stem cells[J]. Digit Libr Forum, 2017(9): 28-34.
[14]	Luciano JS, Andersson B, Batchelor C, et al. The translational medicine ontology and knowledge base: driving personalized medicine by bridging the gap between bench and bedside[J]. J Biomed Semantics, 2011, 2(Suppl 2): S1.
[15]	Lam HYK, Marenco L, Clark T, et al. AlzPharm: integration of neurodegeneration data using RDF[J]. BMC Bioinformatics, 2007, 8(Suppl 3): S4.
[16]	Pearson WR. An introduction to sequence similarity(homology)searching[J]. Curr Protoc Bioinformatics, 2013, Chapter 3: 3.1.1-3.1.8.
[17]	Hu G, Kurgan L. Sequence similarity searching[J]. Curr Protoc Protein Sci, 2019, 95(1): e71. doi: 10.1002/cpps.v95.1 URL
[18]	Launay G, Simonson T. Homology modelling of protein-protein complexes: a simple method and its possibilities and limitations[J]. BMC Bioinformatics, 2008, 9: 427. doi: 10.1186/1471-2105-9-427 pmid: 18844985
[19]	Pearson WR. Selecting the right similarity-scoring matrix[J]. Curr Protoc Bioinformatics, 2013, 43: 3.5.1-3.5.9.
[20]	Wu L, Zhang DF, Xue M, et al. Overexpression of the maize GRF10, an endogenous truncated growth-regulating factor protein, leads to reduction in leaf size and plant height[J]. J Integr Plant Biol, 2014, 56(11): 1053-1063. doi: 10.1111/jipb.v56.11 URL
[21]	Lee SM, Park CM. Regulation of reactive oxygen species generation under drought conditions in Arabidopsis[J]. Plant Signal Behav, 2012, 7(6): 599-601. doi: 10.4161/psb.19940 URL
[22]	Zheng XN, Chen B, Lu GJ, et al. Overexpression of a NAC transcription factor enhances rice drought and salt tolerance[J]. Biochem Biophys Res Commun, 2009, 379(4): 985-989. doi: 10.1016/j.bbrc.2008.12.163 URL
[23]	Zhao X, Wu TT, Guo SX, et al. Ectopic expression of AeNAC83, a NAC transcription factor from Abelmoschus esculentus, inhibits growth and confers tolerance to salt stress in Arabidopsis[J]. Int J Mol Sci, 2022, 23(17): 10182. doi: 10.3390/ijms231710182 URL
[24]	Chen X, Lu SC, Wang YF, et al. OsNAC2 encoding a NAC transcription factor that affects plant height through mediating the gibberellic acid pathway in rice[J]. Plant J, 2015, 82(2): 302-314. doi: 10.1111/tpj.2015.82.issue-2 URL
[25]	Jiang DG, Chen WT, Dong JF, et al. Overexpression of miR164b-resistant OsNAC2 improves plant architecture and grain yield in rice[J]. J Exp Bot, 2018, 69(7): 1533-1543. doi: 10.1093/jxb/ery017 pmid: 29365136
[26]	Zhao H, Li J, Yang L, et al. An inferred functional impact map of genetic variants in rice[J]. Molecular Plant, 2021, 14: 1584-1599 doi: 10.1016/j.molp.2021.06.025 pmid: 34214659
[27]	Zhang P, Wang Y, Chachar S, et al. eRice: A refined epigenomic platform for Japonica and Indica rice[J]. Plant Biotechnology Journal, 2020, 18: 1642-1644. doi: 10.1111/pbi.13329 pmid: 31916375
[28]	Gui S, Yang L, Li J, et al. ZEAMAP, a comprehensive database adapted to the maize multi-omics era[J]. iScience, 2020, 23(6): 101241. doi: 10.1016/j.isci.2020.101241 URL
[29]	Zhang L, Dong C, Chen Z, et al. WheatGmap: A comprehensive platform for wheat gene mapping and genomic studies[J]. Molecular Plant, 2021, 14(2): 187-190. doi: 10.1016/j.molp.2020.11.018 pmid: 33271333