Ontology-aware deep learning enables ultrafast and interpretable source tracking among sub-million microbial community samples from hundreds of niches

Zha, Yuguo; Chong, H. Gin; Qiu, Hao Bo; Kaufmann, Kai; Dun, Yuzheng; Chen, Zhixue; Cui, Xuefeng; Ning, Kang

doi:10.1186/s13073-022-01047-5

Cited by 12 publications

(11 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, when the number of samples and biomes increases, running time increases rapidly, preventing large-scale source tracking. This problem could be solved by deep learning solutions by utilizing model-based methods such as neural networks that would enable improvements in both speed and accuracy during source tracking [61] , [62] .…”

Section: The Dilemma Of Traditional Methods Could Be Solved By Deep L...mentioning

confidence: 99%

“…The random forest approach is more widely used to identify microbial community sources via application toward the prediction of locations and times for forensic studies [7] , [57] , [58] , in addition to application in predicting sources of contamination [59] , [60] . ONN4MST [61] is a deep learning method which employs a neural network model to source track microbial communities at high efficiency and accuracy without any prior knowledge about the microbial communities to be estimated. Its pre-built biome ontology includes 60 environmental biomes, 25 host-associated biomes, and 10 engineered biomes, which represent the most comprehensive potential sources utilized for source tracking.…”

Section: Dark Matter In the Microbiome And The Computational Mining T...mentioning

confidence: 99%

See 1 more Smart Citation

Microbial Dark Matter: From Discovery to Applications

Zha

Chong

Yang

et al. 2022

Genomics, Proteomics &Amp; Bioinformatics

Self Cite

View full text Add to dashboard Cite

With the rapid increase of the microbiome samples and sequencing data, more and more knowledge about microbial communities has been gained. However, there is still much more to learn about microbial communities, including billions of novel species and genes, as well as countless spatiotemporal dynamic patterns within the microbial communities, which together form the microbial dark matter. In this work, we summarized the dark matter in microbiome research and reviewed current data mining methods, especially artificial intelligence (AI) methods, for different types of knowledge discovery from microbial dark matter. We also provided case studies on using AI methods for microbiome data mining and knowledge discovery. In summary, we view microbial dark matter not as a problem to be solved but as an opportunity for AI methods to explore, with the goal of advancing our understanding of microbial communities, as well as developing better solutions to global concerns about human health and the environment.

show abstract

Section: The Dilemma Of Traditional Methods Could Be Solved By Deep L...mentioning

confidence: 99%

Section: Dark Matter In the Microbiome And The Computational Mining T...mentioning

confidence: 99%

Microbial Dark Matter: From Discovery to Applications

Zha

Chong

Yang

et al. 2022

Genomics, Proteomics &Amp; Bioinformatics

Self Cite

View full text Add to dashboard Cite

show abstract

“…Until now, finding suitable fingerprint factors to determine the main dispersal sources of sediment has been a formidable challenge (Vercruysse et al, 2017). Fortunately, some research shows that microorganisms can be used as a special fingerprint factor to trace their host origin due to their genetic diversity, biochemical characteristics and ability to produce a variety of specific metabolites in the environment (Zha et al, 2022). More recently, Zhang et al (2019) concluded that microorganisms can act as helpful indications for sediment sources, based on the indivisible relationship between sediment and microorganisms (Droppo, 2001).…”

Section: Introductionmentioning

confidence: 99%

Assembly process and source tracking of microbial communities in sediments of Dongting Lake

et al. 2023

View full text Add to dashboard Cite

Soil erosion resulted in homogenization of bacterial communities in the watershed.• Microbial community heterogeneity among erosion sites made soil tracing possible.• Assembly process results showed that the tracking results can achieve high precision.• Dryland was the main source of sediment deposition based on the result of FEAST. Sediment source tracing can accurately provide a theoretical basis for controlling soil erosion effectively, by identifying the most serious types of land use. Traditional sediment tracing methods are based on physical, chemical, biological, and composite fingerprinting, which have not included microbes. As high-throughput sequencing becomes more prevalent, microorganisms can provide more information than what we think. Thus, whether the microorganism can also be used as a special fingerprint factor for sediment source identification during soil erosion, we have tested it by using microbial source tracking tool FEAST to quantify the microbe contribution from five types of eroded land (including dryland, urban, paddy field, forest and grassland) to the depositional areas (Niubitan) in the Yuanjiang basin. The source microbial community in the erosive area was heterogeneous, and assembly process analysis further demonstrated that the source tracking results could reach higher accuracy. The results of FEAST showed that dryland (35.50%), urban (17.21%), paddy field (8.14%), and forest (1.07%) were the major contributors to Niubitan. Our results follow the general soil erosion rules and prove its validity. Taken together, a new perspective is provided by these results for tracing sediment sources in erosion-sedimentary systems.

show abstract

“…For instance, performing microbial community classification among thousands of samples within hundreds of biomes may take hours for these methods [ 19 ]. Recently, ONN4MST was proposed to solve the irreconcilable contradiction between efficiency and accuracy [ 20 ]. ONN4MST is a supervised learning method based on ontology-aware neural network (ONN) model, which contains multiple output layers fitting with a general biome ontology.…”

Section: Introductionmentioning

confidence: 99%

EXPERT: transfer learning-enabled context-aware microbial community classification

Chong

Zha

et al. 2022

Briefings in Bioinformatics

Self Cite

View full text Add to dashboard Cite

Microbial community classification enables identification of putative type and source of the microbial community, thus facilitating a better understanding of how the taxonomic and functional structure were developed and maintained. However, previous classification models required a trade-off between speed and accuracy, and faced difficulties to be customized for a variety of contexts, especially less studied contexts. Here, we introduced EXPERT based on transfer learning that enabled the classification model to be adaptable in multiple contexts, with both high efficiency and accuracy. More importantly, we demonstrated that transfer learning can facilitate microbial community classification in diverse contexts, such as classification of microbial communities for multiple diseases with limited number of samples, as well as prediction of the changes in gut microbiome across successive stages of colorectal cancer. Broadly, EXPERT enables accurate and context-aware customized microbial community classification, and potentiates novel microbial knowledge discovery.

show abstract

Ontology-aware deep learning enables ultrafast and interpretable source tracking among sub-million microbial community samples from hundreds of niches

Cited by 12 publications

References 45 publications

Microbial Dark Matter: From Discovery to Applications

Microbial Dark Matter: From Discovery to Applications

Assembly process and source tracking of microbial communities in sediments of Dongting Lake

EXPERT: transfer learning-enabled context-aware microbial community classification

Contact Info

Product

Resources

About