Utilizing machine learning with knockoff filtering to extract significant metabolites in Crohn’s disease with a publicly available untargeted metabolomics dataset

Masud, Shoaib Bin; Jenkins, Conor; Hussey, Erika; Elkin-Frankston, Seth; Mach, Phillip M.; Dhummakupt, Elizabeth S.; Aeron, Shuchin

doi:10.1371/journal.pone.0255240

Cited by 4 publications

(4 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Metabolites with negative weights were hydroxystyrene, eicosatrienoate, and 12,13-dihydroxyoctadec-9-enoic acid (diHOME). 12,13-diHOME was also reported as an important metabolite for differentiating IBD status when applying a knockoff filtering-based multivariate approach to data from HMP2 ( 47 ).…”

Section: Discussionmentioning

confidence: 99%

Poly-omic risk scores predict inflammatory bowel disease diagnosis

Arehart,

Sterrett,

Garris

et al. 2024

mSystems

View full text Add to dashboard Cite

Inflammatory bowel disease (IBD) is characterized by complex etiology and a disrupted colonic ecosystem. We provide a framework for the analysis of multi-omic data, which we apply to study the gut ecosystem in IBD. Specifically, we train and validate models using data on the metagenome, metatranscriptome, virome, and metabolome from the Human Microbiome Project 2 IBD multi-omic database, with 1,785 repeated samples from 130 individuals (103 cases and 27 controls). After splitting the participants into training and testing groups, we used mixed-effects least absolute shrinkage and selection operator regression to select features for each omic. These features, with demographic covariates, were used to generate separate single-omic prediction scores. All four single-omic scores were then combined into a final regression to assess the relative importance of the individual omics and the predictive benefits when considered together. We identified several species, pathways, and metabolites known to be associated with IBD risk, and we explored the connections between data sets. Individually, metabolomic and viromic scores were more predictive than metagenomics or metatranscriptomics, and when all four scores were combined, we predicted disease diagnosis with a Nagelkerke’s R 2 of 0.46 and an area under the curve of 0.80 (95% confidence interval: 0.63, 0.98). Our work supports that some single-omic models for complex traits are more predictive than others, that incorporating multiple omic data sets may improve prediction, and that each omic data type provides a combination of unique and redundant information. This modeling framework can be extended to other complex traits and multi-omic data sets. IMPORTANCE Complex traits are characterized by many biological and environmental factors, such that multi-omic data sets are well-positioned to help us understand their underlying etiologies. We applied a prediction framework across multiple omics (metagenomics, metatranscriptomics, metabolomics, and viromics) from the gut ecosystem to predict inflammatory bowel disease (IBD) diagnosis. The predicted scores from our models highlighted key features and allowed us to compare the relative utility of each omic data set in single-omic versus multi-omic models. Our results emphasized the importance of metabolomics and viromics over metagenomics and metatranscriptomics for predicting IBD status. The greater predictive capability of metabolomics and viromics is likely because these omics serve as markers of lifestyle factors such as diet. This study provides a modeling framework for multi-omic data, and our results show the utility of combining multiple omic data types to disentangle complex disease etiologies and biological signatures.

show abstract

Section: Discussionmentioning

confidence: 99%

Poly-omic risk scores predict inflammatory bowel disease diagnosis

Arehart,

Sterrett,

Garris

et al. 2024

mSystems

View full text Add to dashboard Cite

show abstract

“…12,13-diHOME was also reported as an important metabolite for differentiating IBD status when applying a knockoff filtering-based multivariate approach to data from HMP2. 45…”

Section: Discussionmentioning

confidence: 99%

“…12,13-diHOME was also reported as an important metabolite for differentiating IBD status when applying a knockoff filtering-based multivariate approach to data from HMP2. 45 Our model selected two compounds annotated as taurine that had been isolated using different chromatographic columns (hydrophilic interaction liquid chromatography (HILIC) negative and HILIC positive), and interestingly, these compounds had contrasting positive and negative weights. This may be related to the charge (or other aspects) of the identified compounds, but more detailed investigation to follow up the untargeted metabolomics would be needed to differentiate these two.…”

Section: Interpretation Of Features Identified By Individual Modelsmentioning

confidence: 99%

Poly-omic risk scores predict inflammatory bowel disease diagnosis

Arehart

Sterrett

Garris

et al. 2022

Preprint

View full text Add to dashboard Cite

Inflammatory Bowel Disease (IBD) is characterized by complex etiology and a disrupted gut microbiota. The substantial non-genetic variance for Crohn's disease and ulcerative colitis (≥25%) suggests that both genetic and environmental factors contribute to IBD development. We aim to extend the framework of genomic studies by examining gut characteristics that are affected by both genetic and environmental factors. Specifically, we train models and validate their accuracy using data that quantifies the microbiota, their transcripts, and the metabolites present in the gut. The IBD Multi-omics Database from the Human Microbiome Project 2 provided 1,785 repeated samples for 131 individuals (103 cases, 27 controls) across multiple -omics layers including metagenomics, metatranscriptomics, viromics, and metabolomics. After splitting the subjects into training and validation groups, we used mixed effects least absolute shrinkage and selection operator (LASSO) regression to determine the most relevant features for each -omic layer. These features, along with demographic covariates, were incorporated into a polygenic risk score framework to generate four separate -omic-level prediction scores. All four -omic-level scores were then combined into a final regression to assess the relative importance of individual -omics and the added benefits when considered together. Our models identified several species, pathways, and metabolites known to be associated with IBD risk. Individually, metabolomics and viromics based scores were more predictive than metagenomics or metatranscriptomics based scores, and when all four scores were combined, we predicted disease diagnosis with a Nagelkerke's R2 of 0.46 and an AUC of 0.80 [95% CI: 0.63, 0.98].

show abstract

“…Machine learning (ML), a subfield of artificial intelligence, has the ability to recognize patterns in large data environments from untargeted metabolomics in a way the human mind is not trained and is able to make decisions or predictions . Thus, through the integration of mass spectrometry and machine learning, several platforms have been built for developing diagnostics tools using experimental metabolomics. − …”

Section: Introductionmentioning

confidence: 99%

Metabolomic and Lipidomic Characterization of Meningioma Grades Using LC–HRMS and Machine Learning

Safari Yazd,

Bazargani,

Fitzpatrick

et al. 2023

J. Am. Soc. Mass Spectrom.

View full text Add to dashboard Cite

Meningiomas are among the most common brain tumors that arise from the leptomeningeal cover of the brain and spinal cord and account for around 37% of all central nervous system tumors. According to the World Health Organization, meningiomas are classified into three histological subtypes: benign, atypical, and anaplastic. Sometimes, meningiomas with a histological diagnosis of benign tumors show clinical characteristics and behavior of aggressive tumors. In this study, we examined the metabolomic and lipidomic profiles of meningioma tumors, focusing on comparing low-grade and high-grade tumors and identifying potential markers that can discriminate between benign and malignant tumors. High-resolution mass spectrometry coupled to liquid chromatography was used for untargeted metabolomics and lipidomics analyses of 85 tumor biopsy samples with different meningioma grades. We then applied feature selection and machine learning techniques to find the features with the highest information to aid in the diagnosis of meningioma grades. Three biomarkers were identified to differentiate low-and high-grade meningioma brain tumors. The use of mass-spectrometry-based metabolomics and lipidomics combined with machine learning analyses to prospect and characterize biomarkers associated with meningioma grades may pave the way for elucidating potential therapeutic and prognostic targets.

show abstract

Utilizing machine learning with knockoff filtering to extract significant metabolites in Crohn’s disease with a publicly available untargeted metabolomics dataset

Cited by 4 publications

References 48 publications

Poly-omic risk scores predict inflammatory bowel disease diagnosis

Poly-omic risk scores predict inflammatory bowel disease diagnosis

Poly-omic risk scores predict inflammatory bowel disease diagnosis

Metabolomic and Lipidomic Characterization of Meningioma Grades Using LC–HRMS and Machine Learning

Contact Info

Product

Resources

About