Analysis of Gut Microbiome Using Explainable Machine Learning Predicts Risk of Diarrhea Associated With Tyrosine Kinase Inhibitor Neratinib: A Pilot Study

Wong, Chi Wah; Yost, Susan E.; Lee, Jin Sung; Gillece, John D.; Folkerts, Megan; Reining, Lauren; Highlander, Sarah K.; Eftekhari, Zahra; Mortimer, Joanne; Yuan, Yuan

doi:10.3389/fonc.2021.604584

Cited by 20 publications

(16 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Pharmacomicrobiomics focuses on the interplay between gut microbiota and drug response through alterations in pharmacokinetics (ie, modification of drug absorption, distribution, metabolism or elimination) or pharmacodynamics (ie, modification of drug targets or biological pathways leading to varying sensitivity to pharmacological effects). 12 Growing evidence has revealed the intimate relationship between the gut microbiota and anticancer treatments, including chemotherapy, 14 radiotherapy, 15 targeted therapy 16 and immunotherapy. 1 Harnessing the microbiota to optimise cancer treatments has become an alternate avenue for personalised medicine.…”

Section: Introductionmentioning

confidence: 99%

Cancer pharmacomicrobiomics: targeting microbiota to optimise cancer therapy outcomes

Ting

Lau

2022

Gut

143

View full text Add to dashboard Cite

Despite the promising advances in novel cancer therapy such as immune checkpoint inhibitors (ICIs), limitations including therapeutic resistance and toxicity remain. In recent years, the relationship between gut microbiota and cancer has been extensively studied. Accumulating evidence reveals the role of microbiota in defining cancer therapeutic efficacy and toxicity. Unlike host genetics, microbiota can be easily modified via multiple strategies, including faecal microbiota transplantation (FMT), probiotics and antibiotics. Preclinical studies have identified the mechanisms on how microbes influence cancer treatment outcomes. Clinical trials have also demonstrated the potential of microbiota modulation in cancer treatments. Herein, we review the mechanistic insights of gut microbial interactions with chemotherapy and ICIs, particularly focusing on the interplay between gut bacteria and the pharmacokinetics (eg, metabolism, enzymatic degradation) or pharmacodynamics (eg, immunomodulation) of cancer treatment. The translational potential of basic findings in clinical settings is then explored, including using microbes as predictive biomarkers and microbial modulation by antibiotics, probiotics, prebiotics, dietary modulations and FMT. We further discuss the current limitations of gut microbiota modulation in patients with cancer and suggest essential directions for future study. In the era of personalised medicine, it is crucial to understand the microbiota and its interactions with cancer. Manipulating the gut microbiota to augment cancer therapeutic responses can provide new insights into cancer treatment.

show abstract

Section: Introductionmentioning

confidence: 99%

Cancer pharmacomicrobiomics: targeting microbiota to optimise cancer therapy outcomes

Ting

Lau

2022

Gut

143

View full text Add to dashboard Cite

show abstract

“…Fourth, endoR is flexible: it can be applied to both random forests and gradient boosted trees, which themselves can be applied to both regression and classification tasks involving various types of features (e.g., microbial abundances and metadata). Finally, endoR is substantially more computationally efficient than, for example, SHAP (Supplementary Figure 8 A-D, and G), which is a common approach for ML model interpretation (31, 35, 92, 93).…”

Section: Discussionmentioning

confidence: 99%

Interpreting tree ensemble machine learning models with endoR

Ruaud¹,

Pfister

Ley³

et al. 2022

Preprint

View full text Add to dashboard Cite

Background: Tree ensemble machine learning models are increasingly used in microbiome science as they are compatible with the compositional, high-dimensional, and sparse structure of sequence-based microbiome data. While such models are often good at predicting phenotypes based on microbiome data, they only yield limited insights into how microbial taxa or genomic content may be associated. Results: We developed endoR, a method to interpret a fitted tree ensemble model. First, endoR simplifies the fitted model into a decision ensemble from which it then extracts information on the importance of individual features and their pairwise interactions and also visualizes these data as an interpretable network. Both the network and importance scores derived from endoR provide insights into how features, and interactions between them, contribute to the predictive performance of the fitted model. Adjustable regularization and bootstrapping help reduce the complexity and ensure that only essential parts of the model are retained. We assessed the performance of endoR on both simulated and real metagenomic data. We found endoR to infer true associations with more or comparable accuracy than other commonly used approaches while easing and enhancing model interpretation. Using endoR, we also confirmed published results on gut microbiome differences between cirrhotic and healthy individuals. Finally, we utilized endoR to gain insights into components of the microbiome that predict the presence of human gut methanogens, as these hydrogen-consumers are expected to interact with fermenting bacteria in a complex syntrophic network. Specifically, we analyzed a global metagenome dataset of 2203 individuals and confirmed the previously reported association between Methanobacteriaceae and Christensenellales. Additionally, we observed that Methanobacteriaceae are associated with a network of hydrogen-producing bacteria. Conclusion: Our method accurately captures how tree ensembles use features and interactions between them to predict a response. As demonstrated by our applications, the resultant visualizations and summary outputs facilitate model interpretation and enable the generation of novel hypotheses about complex systems. An implementation of endoR is available as an open-source R-package on GitHub (https://github.com/leylabmpi/endoR).

show abstract

“…In turn, a growing number of studies have demonstrated the benefit of leveraging machine learning to differentiate sample classes in microbiome studies. These include predicting the risk of type 2 diabetes (19), diarrhea associated with cancer treatment (20), and liver disease (21) from the gut microbiome. Additionally, when sampling the microbiome from leg skin, explainable AI was able to identify microbial drivers associated with skin hydration, age, and pre/post-menopausal status from the skin microbiome (22).…”

Section: Introductionmentioning

confidence: 99%

Development and evaluation of statistical and Artificial Intelligence approaches with microbial shotgun metagenomics data as an untargeted screening tool for use in food production

Haiminen¹,

Beck

Agarwal

et al. 2022

Preprint

View full text Add to dashboard Cite

The increasing knowledge of microbial ecology in food products relating to quality and safety as well as the established usefulness of machine learning algorithms for anomaly detection in multiple scenarios suggests that the application of microbiome data in food production systems for anomaly detection is an important next step toward bettering our food supply chain. To that end, the objective of the research presented here was to assess the feasibility of using whole metagenome shotgun sequencing data as input into anomaly detection algorithms using fluid milk as a model system. Contrastive PCA, cluster-based methods, and explainable AI were evaluated as potential methods for the detection of two anomalous sample classes using longitudinal metagenomic profiling of fluid milk compared to baseline samples collected under comparable circumstances. Traditional methods (alpha and beta diversity, clustering based contrastive PCA, MDS, and dendrograms) failed to differentiate anomalous sample classes; however, explainable AI was able to classify anomalous vs. baseline samples and indicate microbial drivers in association with antibiotic use e.g., Morganella and Enterobacter and microbial signatures related to milk collected from an alternative farm e.g., Coxiella. This work also characterizes the microbiome of fluid milk in greater depth than previously published studies. Our results indicate the application of artificial intelligence continues to hold promise in the realm of microbiome data analysis and could present further opportunity for downstream analytic automation to aid in food safety and quality.

show abstract

Analysis of Gut Microbiome Using Explainable Machine Learning Predicts Risk of Diarrhea Associated With Tyrosine Kinase Inhibitor Neratinib: A Pilot Study

Cited by 20 publications

References 39 publications

Cancer pharmacomicrobiomics: targeting microbiota to optimise cancer therapy outcomes

Cancer pharmacomicrobiomics: targeting microbiota to optimise cancer therapy outcomes

Interpreting tree ensemble machine learning models with endoR

Development and evaluation of statistical and Artificial Intelligence approaches with microbial shotgun metagenomics data as an untargeted screening tool for use in food production

Contact Info

Product

Resources

About