Human disease prediction from microbiome data by multiple feature fusion and deep learning

Chen, Xingjian; Zhu, Zifan; Zhang, Weitong; Wang, Yuchen; Wong, Ka-Chun

doi:10.1016/j.isci.2022.104081

Cited by 15 publications

(22 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We first used the same six datasets as those used for developing ML models including MetAML [19], DeepMicro [6] and EPCNN [9]. Table 1 summarizes the six datasets and the diseases they represent.…”

Section: Collection Of the Human Gut Metagenomic Samplesmentioning

confidence: 99%

“…b : based on the results we derived by running DeepMicro on the same input we prepared for all included approaches. c : the AUCs for EPCNN were taken from [9]. Note the reported AUCs by EPCNN were based on predictors also trained using species abundances information, but the difference is that it used both known and unknown species, and the quantification approach was different.…”

Section: Comparison Of Microkpnn With Fully-connected Nn and Existing...mentioning

confidence: 99%

“…PopPhy-CNN prepares the 2D matrix representing the phylogenetic tree populated with the relative abundance of microbial taxa in a metagenomic sample, such that, for a given row, the children of the nodes from that row are selected and their abundances are placed in the subsequent row in the order that their parents appear, starting with the left-most column. EPCNN [9] is a more recent effort that uses an ensemble strategy utilizing different microbial features and taxonomic representation for microbiome-based prediction, aiming to reduce overfitting.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Incorporating metabolic activity, taxonomy and community structure to improve microbiome-based predictive models for host phenotype prediction

Monshizadeh

2023

Preprint

View full text Add to dashboard Cite

We developed MicroKPNN, a prior-knowledge guided interpretable neural network for microbiome-based human host phenotype prediction. The prior-knowledge used in MicroKPNN includes the metabolic activities of different bacterial species, phylogenetic relationships, and bacterial community structure. Application of MicroKPNN to seven gut microbiome datasets (involving five different human diseases including inflammatory bowel disease, type 2 diabetes, liver cirrhosis, colorectal cancer, and obesity) shows that incorporation of the prior knowledge helped improve the microbiome-based host phenotype prediction. MicroKPNN outperformed fully-connected neural network based approaches in all seven cases, with the most improvement of accuracy in the prediction of type 2 diabetes. MicroKPNN outperformed a recently developed deep-learning based approach DeepMicro, which selects the best combination of autoencoder and machine learning approach to make predictions, in six out of the seven cases. More importantly, we showed that MicroKPNN provides a way for interpretation of the predictive models. Our results suggested that the metabolic potential of the bacterial species contributed more than the two other sources of prior knowledge. MicroKPNN is publicly available at https://github.com/mgtools/MicroKPNN. Keywords: gut microbiome, human disease, interpretable neural network, prior-knowledge primed, metabolic activity, taxonomy, bacterial community

show abstract

Section: Collection Of the Human Gut Metagenomic Samplesmentioning

confidence: 99%

Section: Comparison Of Microkpnn With Fully-connected Nn and Existing...mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Incorporating metabolic activity, taxonomy and community structure to improve microbiome-based predictive models for host phenotype prediction

Monshizadeh

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Finally, another aspect that can be taken into account when taxonomy is studied is the fact that a great part of it is unknown, whether it is because abundance is obtained by unsupervised binning or because reads come from unknown species. MetaDR ( [143]) takes into account both known and unknown features as well as the topology of the taxonomy tree obtained by converting it to an image, allowing MetaDR to compete with the best state-of-the-art methods, while showing good computational speed and ranking among the best taxonomy-based methods.…”

Section: Sequence-based Approachesmentioning

confidence: 99%

Deep learning methods in metagenomics: a review

Roy

Prifti

Zucker

2023

Preprint

View full text Add to dashboard Cite

The ever decreasing cost of sequencing and the multiplication of potential applications for the study of metagenomes have led to an unprecedented increase in the volume of data generated. One of the most prevalent applications of metagenomics is the study of microbial environments, such as the human gut. The gut microbiome has been shown to play an important role in human health, providing critical information for patient diagnosis and prognosis. However, the analysis of metagenomic data remains challenging for many reasons, including reference catalogs, sparsity and compositionality of the data, to name a few. Deep learning (DL) enables novel and promising approaches that complement state-of-the-art microbiome pipelines. In fact, DL-based methods can address almost all aspects of microbiome analysis, including novel pathogen detection, sequence classification, patient stratification, and disease prediction. Beyond the generation of predictive models, a key aspect of such methods remains their interpretability. In this article, we provide a systematic review of deep learning approaches in metagenomics, whether based on convolutional networks, autoencoders, or attention-based models. These methods aggregate contextualized data and pave the way for improved patient care and a better understanding of the key role the microbiome plays in our health.

show abstract

“…Their results show that the DNN-based framework accelerates the model training process and improves the model performance of disease prediction. Moreover, several DNN-based architectures integrating with the phylogenetic tree, such as EPCNN [ 136 ], Ph-CNN [ 140 ], PopPhy-CNN [ 141 ], have been developed, which show an improved performance.…”

Section: Application Of Machine Learning For Analysis Of Gut Microbio...mentioning

confidence: 99%

Machine learning for data integration in human gut microbiome

Luo

et al. 2022

Microb Cell Fact

View full text Add to dashboard Cite

Recent studies have demonstrated that gut microbiota plays critical roles in various human diseases. High-throughput technology has been widely applied to characterize the microbial ecosystems, which led to an explosion of different types of molecular profiling data, such as metagenomics, metatranscriptomics and metabolomics. For analysis of such data, machine learning algorithms have shown to be useful for identifying key molecular signatures, discovering potential patient stratifications, and particularly for generating models that can accurately predict phenotypes. In this review, we first discuss how dysbiosis of the intestinal microbiota is linked to human disease development and how potential modulation strategies of the gut microbial ecosystem can be used for disease treatment. In addition, we introduce categories and workflows of different machine learning approaches, and how they can be used to perform integrative analysis of multi-omics data. Finally, we review advances of machine learning in gut microbiome applications and discuss related challenges. Based on this we conclude that machine learning is very well suited for analysis of gut microbiome and that these approaches can be useful for development of gut microbe-targeted therapies, which ultimately can help in achieving personalized and precision medicine.

show abstract

Human disease prediction from microbiome data by multiple feature fusion and deep learning

Cited by 15 publications

References 44 publications

Incorporating metabolic activity, taxonomy and community structure to improve microbiome-based predictive models for host phenotype prediction

Incorporating metabolic activity, taxonomy and community structure to improve microbiome-based predictive models for host phenotype prediction

Deep learning methods in metagenomics: a review

Machine learning for data integration in human gut microbiome

Contact Info

Product

Resources

About