Content Estimating of Microbial Dissolved Organic Carbon Based on Machine Learning

doi:10.19586/j.2095-2341.2022.0007

Current Biotechnology ›› 2023, Vol. 13 ›› Issue (4): 645-653.DOI: 10.19586/j.2095-2341.2022.0007

• Techniques and Methods • Previous Articles

Content Estimating of Microbial Dissolved Organic Carbon Based on Machine Learning

Yunpeng MA(), Jing ZHU(), Xinghua CUI

College of Computer and Information Engineering，Xinjiang Agricultural University，Urumqi 830052，China

Received:2022-01-17 Accepted:2022-03-09 Online:2023-07-25 Published:2023-08-03
Contact: Jing ZHU

基于机器学习的微生物溶解有机碳含量估测

马云鹏(), 朱静(), 崔兴华

新疆农业大学计算机与信息工程学院，乌鲁木齐 830052

通讯作者: 朱静
作者简介:马云鹏 E-mail：1020264950@qq.com；
基金资助:
新疆畜牧科学院畜牧研究所基础研究项目(2020BD1002-2-2-2)

Abstract

Abstract:

The microbial communities has an important impact on the macro nature of the environment. However， the characteristics of high-dimensional， complex and sparse microbial data also pose new challenges for understanding the relationship between microorganisms and ecological environment. The development of machine learning and the popularization of the application of the second generation DNA sequencing technology provided a new solution to this problem. In this study， soil microbiome and dissolved organic carbon （DOC） data of 308 samples from plant litter decomposition experiments for 44 days were used， and 1 709 operational taxonomic units （OTU） of bacteria and microorganisms were used as features to build 12 commonly used machine learning models. Embedding method， packaging method and embedd-packaging fusion method were used for feature selection， and gradient boosting decision tree （GBDT） was selected as the optimal model for parameter optimization. The model adopted root mean square error， mean absolute error and linear goodness of fit was used as evaluation indexes. The results showed that， the feature selection reduced the data dimension and improved the model accuracy. In the simulation experiment， the embedding-packaging fusion method performs was the best in the application model. The prediction model of dissolved organic carbon was constructed by combining the embedding and packaging fusion method with gradient boosting decision tree， and the validity of the model was verified by experiments. The results provided a new way to estimate dissolved organic carbon using machine learning method based on bacterial and microbial data.

Key words: machine learning, microorganism, feature screening, modeling prediction, organic carbon

摘要：

微生物群落会对所处环境的宏观性质产生重要影响，但微生物存在数据高维、复杂、稀疏的特点，为了解微生物与生态环境之间的关系提出了新的挑战。机器学习的发展以及第二代DNA测序技术应用的普及为解决这一问题提供了一种新的方法。利用308个样本共44 d的植物凋落物分解实验的土壤微生物群和溶解有机碳（dissolved organic carbon，DOC）数据，并以1 709个细菌微生物操作分类单元（operational taxonomic units，OTU）作为特征构建12种常用的机器学习模型，采用嵌入法、包装法以及嵌入-包装融合法进行特征选择，并选择梯度提升决策树（gradient boosting decision tree，GBDT）作为最优模型进行参数优化。模型采用均方根误差、平均绝对误差、线性拟合优度作为评价指标。结果表明，特征筛选后降低了数据维度，提升了模型精度，在仿真实验中，嵌入-包装融合法在应用模型中表现最佳。将嵌入-包装融合法与梯度提升决策树结合构建溶解有机碳预测模型，并通过实验验证了模型的有效性。研究结果为利用细菌微生物数据应用机器学习方法估测溶解有机碳提供了新思路。

关键词: 机器学习, 微生物, 特征筛选, 建模预测, 有机碳

CLC Number:

TP181

Yunpeng MA, Jing ZHU, Xinghua CUI. Content Estimating of Microbial Dissolved Organic Carbon Based on Machine Learning[J]. Current Biotechnology, 2023, 13(4): 645-653.

马云鹏, 朱静, 崔兴华. 基于机器学习的微生物溶解有机碳含量估测[J]. 生物技术进展, 2023, 13(4): 645-653.

Figures/Tables 16

References 19

1	LI H Z. Microbiome, metagenomics, and high-dimensional compositional data analysis[J]. Ann. Rev. Stat. Appl., 2015, 2: 73-94.
2	SARKER I H. Machine learning: Algorithms, real-world applications and research directions[J]. SN Comput. Sci., 2021, 2(3): 1-21.
3	HASAN B M S, ABDULAZEEZ A M. A review of principal component analysis algorithm for dimensionality reduction[J]. J. Soft Comput. Data Min., 2021, 2(1): 20-30.
4	STATNIKOV A, HENAFF M, NARENDRA V, et al.. A comprehensive evaluation of multicategory classification methods for microbiomic data[J]. Microbiome, 2013, 1(1): 1-12.
5	ZELLER G, TAP J, VOIGT A Y, et al.. Potential of fecal microbiota for early:tage detection of colorectal cancer[J/OL]. Mol. Syst. Biol., 2014, 10(11): 766[2022-05-06]. .
6	NING J, BEIKO R G. Phylogenetic approaches to microbial community classification[J]. Microbiome, 2015, 3(1): 1-13.
7	LO C, MARCULESCU R. MetaNN: accurate classification of host phenotypes from metagenomic data using neural networks[J]. BMC Bioinform., 2019, 20(12): 1-14.
8	BOKULICH N A, DILLON M R, BOLYEN E, et al.. q2 -sample-classifier: machine-learning tools for microbiome classification and regression[J/OL]. J. Open Res. Softw., 2018, 3(30):934[2022-05-06]. .
9	黄荣才,高胜涛,范士杰,等.畜禽粪污源抗生素及耐药基因在环境中的归趋[J].生物技术进展,2019,9(2):146-151.
10	刘超,王宪伟,宋艳宇,等.增温对冻土区泥炭沼泽土壤孔隙水甲烷关联微生物和溶解性有机碳的影响[J].生态学报,2021,41(1):184-193.
11	丁咸庆,柏菁,项文化,等.不同浸提剂处理森林土壤溶解性有机碳含量比较[J].土壤,2020,52(3):518-524.
12	郭利娜,贾羽旋,李彤,等.森林溶解性有机碳淋溶驱动机制及模拟研究进展[J].生态学杂志,2020,39(5):1723-1733.
13	余高, 陈芬, 谢英荷,等.有机肥替代化肥比例对黄壤土活性有机碳及酶活性的影响[J].中国蔬菜,2020(4):48-55.
14	LIANG C, SCHIMEL J P, JASTROW J D. The importance of anabolism in microbial control over soil carbon storage[J]. Nat. Microbiol., 2017, 2(8): 1-6.
15	ZITNIK M, NGUYEN F, WANG B, et al.. Machine learning for integrating data in biology and medicine: principles, practice, and opportunities[J]. Inf. Fusion, 2019, 50: 71-91.
16	JOHANSEN R, ALBRIGHT M, LOPEZ D, et al.. Microbial community-level features linked to divergent carbon flows during early litter decomposition in a constant environment[J/OL]. BioRxiv, 2019: 659383[2022-05-16]. .
17	AFENDRAS G, MARKATOU M. Optimality of training/test size and resampling effectiveness in cross-validation[J]. J. Stat. Plan. Infer., 2019, 199: 286-301.
18	THOMPSON J, JOHANSEN R, DUNBAR J, et al.. Machine learning to predict microbial community functions: an analysis of dissolved organic carbon from litter decomposition[J/OL]. PLoS ONE, 2019, 14(7): e0215502[2022-05-16]. .
19	SEONWOO M, BYUNGHAN L, SUNGROH Y. Deep learning in bioinformatics[J]. Brief Bioinform., 2017, 18(5): 851-869.

样本	操作分类单元编号							DOC含量/（mg·g^-1）
样本	OTU_401	OTU_12	OTU_960	OTU_20	OTU_11	OTU_6	OTU_25	DOC含量/（mg·g^-1）
样本1	0	38	1	17	0	27	0	10.46
样本2	140	28	0	0	0	3	0	7.00
样本3	262	8	109	11	0	67	18	5.00
样本4	102	9	0	1	0	2	0	8.42
样本5	2	26	6	73	0	0	0	9.46

样本	操作分类单元编号							DOC含量/（mg·g^-1）
样本	OTU_401	OTU_12	OTU_960	OTU_20	OTU_11	OTU_6	OTU_25	DOC含量/（mg·g^-1）
样本1	0	38	1	17	0	27	0	10.46
样本2	140	28	0	0	0	3	0	7.00
样本3	262	8	109	11	0	67	18	5.00
样本4	102	9	0	1	0	2	0	8.42
样本5	2	26	6	73	0	0	0	9.46

算法	RMSE	MAE	R²
套索回归	2.460 6	1.979 1	0.305 2
弹性网回归	2.305 8	1.804 4	0.390 5
支持向量机	2.389 3	1.894 2	0.342 1
决策树	2.779 2	2.131 4	0.065 0
K近邻	2.748 6	2.097 0	0.146 7
多层感知机	2.549 0	2.116 5	0.262 2
极限树	1.995 5	1.551 8	0.539 9
极限梯度提升决策树	2.049 1	1.576 0	0.508 1
随机森林	1.979 1	1.515 4	0.544 5
自适应增强算法	2.073 4	1.620 5	0.503 1
引导聚集算法	2.103 5	1.622 9	0.477 2
梯度提升决策树	1.955 4	1.472 4	0.558 5

算法	RMSE	MAE	R²
套索回归	2.460 6	1.979 1	0.305 2
弹性网回归	2.305 8	1.804 4	0.390 5
支持向量机	2.389 3	1.894 2	0.342 1
决策树	2.779 2	2.131 4	0.065 0
K近邻	2.748 6	2.097 0	0.146 7
多层感知机	2.549 0	2.116 5	0.262 2
极限树	1.995 5	1.551 8	0.539 9
极限梯度提升决策树	2.049 1	1.576 0	0.508 1
随机森林	1.979 1	1.515 4	0.544 5
自适应增强算法	2.073 4	1.620 5	0.503 1
引导聚集算法	2.103 5	1.622 9	0.477 2
梯度提升决策树	1.955 4	1.472 4	0.558 5

操作分类单元编号
OTU_401	OTU_12	OTU_960	OTU_20	OTU_11	OTU_6	OTU_40	OTU_53	OTU_150	OTU_55
OTU_57	OTU_202	OTU_160	OTU_574	OTU_249	OTU_95	OTU_188	OTU_221	OTU_1469	OTU_3824
OTU_273	OTU_389	OTU_292	OTU_636	OTU_23	OTU_101	OTU_4022	OTU_539	OTU_61	OTU_146
OTU_181	OTU_100	OTU_120	OTU_81	OTU_262	OTU_1259	OTU_616	OTU_5	OTU_1019	OTU_1032
OTU_138	OTU_16	OTU_167	OTU_170	OTU_1914	OTU_21	OTU_22	OTU_220	OTU_227	OTU_24
OTU_267	OTU_29	OTU_309	OTU_313	OTU_329	OTU_3858	OTU_473	OTU_474	OTU_534	OTU_597
OTU_70	OTU_82	OTU_98	OTU_1	OTU_15	OTU_8	OTU_7	OTU_13	OTU_18	OTU_26
OTU_10	OTU_9	OTU_45	OTU_1033	OTU_44	OTU_193	OTU_35	OTU_32	OTU_27	OTU_131
OTU_28	OTU_51	OTU_84	OTU_5179	OTU_56	OTU_54	OTU_77	OTU_75	OTU_94	OTU_1297
OTU_1111	OTU_1200	OTU_1974	OTU_103	OTU_2139	OTU_950	OTU_106	OTU_235	OTU_251	OTU_431
OTU_358	OTU_2512	OTU_713	OTU_3826	OTU_179	OTU_211	OTU_1119	OTU_1569	OTU_201	OTU_2586
OTU_3002	OTU_320	OTU_953	OTU_1509	OTU_226	OTU_347	OTU_169	OTU_470	OTU_293	OTU_5841
OTU_363	OTU_357	OTU_407	OTU_458	OTU_372	OTU_1052	OTU_581	OTU_652	OTU_5988	OTU_1550
OTU_545	OTU_698	OTU_1348	OTU_5531	OTU_4794	OTU_2669	OTU_516	OTU_994	OTU_4277	OTU_5059

Content Estimating of Microbial Dissolved Organic Carbon Based on Machine Learning

基于机器学习的微生物溶解有机碳含量估测

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 16

References 19

Related Articles 15

Recommended Articles

Metrics

Comments

特征选择方法	算法	RMSE	MAE	R²
RFE（RF）	梯度提升决策树	1.940 7	1.478 6	0.579 2
	极限树	1.963 1	1.501 0	0.566 9
	随机森林	1.958 5	1.486 0	0.566 9
RFE（GBDT）	梯度提升决策树	1.821 2	1.377 5	0.618 3
	极限树	1.855 1	1.420 5	0.601 5
	随机森林	1.905 8	1.453 7	0.581 8
RFE（ET）	梯度提升决策树	1.954 3	1.487 6	0.556 3
	极限树	1.874 0	1.425 9	0.597 6
	随机森林	1.936 5	1.474 2	0.566 1
FIS（GBDT）	梯度提升决策树	1.864 4	1.412 1	0.601 3
	极限树	1.937 1	1.499 3	0.566 7
	随机森林	1.956 4	1.493 8	0.555 8
RFE-FIS（GBDT）	梯度提升决策树	1.818 8	1.386 8	0.620 3
	极限树	1.914 6	1.466 3	0.577 8
	随机森林	1.924 7	1.459 3	0.570 2

参数	搜索范围	搜索步长
learning_rate	0.01~0.2	0.01
n_estimators	100~1 000	1.00
max_depth	1~10	1.00

[1]	Xiaoqi WU, Wenjing GONG, Guoyu LI, Ang LI, Jihua WANG, Di CUI. Knowledge Gaps and Chanllenges in Microbial Fermentation of Traditional Chinese Medicine: From Strain Selection to Quality Control [J]. Current Biotechnology, 2025, 15(2): 201-211.
[2]	Huanzhen WU, Ye YANG, Xiuming CUI, Yuan LIU. The Current Status and Improvement Strategies of Agricultural Biological Control Technology [J]. Current Biotechnology, 2024, 14(5): 697-711.
[3]	Bicong WU, Bo JIAO, Yu ZHANG, Xin GUO, Yu ZHANG, Xiaohong LUO, Lei DAI, Qiang WANG. Effect of Feed-to-liquid Ratio on the Quality Characteristics of Stirred UHT Walnut Yogurt [J]. Current Biotechnology, 2024, 14(4): 640-648.
[4]	Chang XU, Tianyi LIU, Wenjia LIU, Limin ZHANG, Jixian MO. Research Progress in Source, Biosynthesis and Function of Microbial Exopolysaccharides [J]. Current Biotechnology, 2024, 14(3): 368-376.
[5]	Jiaqi SUN, Jia GUO, Chuang ZHANG, Qing LIU, Ziyu WANG, Hanchao XIA, Buxuan QIAN, Fangfang ZHAO, Qi WANG, Jianfeng LIU, Xiangguo LIU. Research Progress of Phosphite Dehydrogenase in Genetically Engineered Microorganisms and Plants [J]. Current Biotechnology, 2024, 14(2): 173-181.
[6]	Ting XU, Jiahao SHEN, Kang ZHAO, Lu HUANG, Enhui DONG, Kexin ZENG, Xinwei BIAN, Minghui JI, Qin XU. Bacterial Signature for Prediction of Disease Type Based on Abundance of Ruminococcus [J]. Current Biotechnology, 2024, 14(2): 323-330.
[7]	Haitao CAO, Jing ZHU, Yunpeng MA, Xinghua CUI. Application of Machine Learning in Phenotypic Prediction of Gut Microbiota [J]. Current Biotechnology, 2023, 13(5): 671-680.
[8]	Haitao CAO, Jing ZHU, Haibo ZENG, Yanchen LIU. Research on Feature Selection of Gut Microbiota and Disease Prediction Model Based on Weighted Average [J]. Current Biotechnology, 2023, 13(5): 798-806.
[9]	Siyuan QIU, Jingxue XU, Yuyang DUAN, Jinyu ZHAO, Wenjing ZHAO, Lixin ZHANG, Guoling REN. Research Progress on Production and Application of Mannosylerythritol Lipids [J]. Current Biotechnology, 2023, 13(2): 210-219.
[10]	Ruiju MIAO, Zundan DING, Jian TIAN, Hongbing ZHANG, Feifei GUAN. Research Advances on Traditional and Intelligent Molecular Design of PET Hydrolases [J]. Current Biotechnology, 2023, 13(1): 46-54.
[11]	Jie HAO, Qiang JI, Liqun LI, Chao ZHENG, Na WU, Han WU, Xuanwen LI, Zhikang SUN. Research Progress on Improving Aroma of Tobacco Leaf with Bio-enzymes and Microorganism Technology [J]. Current Biotechnology, 2022, 12(6): 817-824.
[12]	Peimin LIU, Jinping LUO, Quanxin GAO. Research Progress of Environmental Microorganisms in Aquaculture [J]. Current Biotechnology, 2022, 12(5): 690-695.
[13]	Zhiqi XIN, Hang ZHAO, Hai WANG, Tiegang LU. Crop Genomics and Genetic Improvement Based on Deep Learning [J]. Current Biotechnology, 2021, 11(4): 483-488.
[14]	XU Huanhuan, ZHANG Hongbing^*, LI Huixuan, LI Lei. Application Progress of Atmospheric and Room Temperature Plasma Technology in Microbial Mutagenesis [J]. Curr. Biotech., 2020, 10(4): 358-362.
[15]	LI Yubang1, WU Junlin1,2*, LI Mansha1. Progress on Microbial Fermentation of Medicinal and Edible Plants [J]. Curr. Biotech., 2019, 9(5): 461-466.