Gut microbiome identifies risk for colorectal polyps

A review of machine learning methods for cancer characterization from microbiome data

Recent studies have shown that the microbiome can impact cancer development, progression, and response to therapies suggesting microbiome-based approaches for cancer characterization. As cancer-related signatures are complex and implicate many taxa, their discovery often requires Machine Learning approaches. This review discusses Machine Learning methods for cancer characterization from microbiome data. It focuses on the implications of choices undertaken during sample collection, feature selection and pre-processing. It also discusses ML model selection, guiding how to choose an ML model, and model validation. Finally, it enumerates current limitations and how these may be surpassed. Proposed methods, often based on Random Forests, show promising results, however insufficient for widespread clinical usage. Studies often report conflicting results mainly due to ML models with poor generalizability. We expect that evaluating models with expanded, hold-out datasets, removing technical artifacts, exploring representations of the microbiome other than taxonomical profiles, leveraging advances in deep learning, and developing ML models better adapted to the characteristics of microbiome data will improve the performance and generalizability of models and enable their usage in the clinic.

Identification of taxonomic changes in the fecal bacteriome associated with colorectal polyps and cancer: potential biomarkers for early diagnosis

Colorectal cancer (CRC) commonly arises in individuals with premalignant colon lesions known as polyps, with both conditions being influenced by gut microbiota. Host-related factors and inherent characteristics of polyps and tumors may contribute to microbiome variability, potentially acting as confounding factors in the discovery of taxonomic biomarkers for both conditions. In this study we employed shotgun metagenomics to analyze the taxonomic diversity of bacteria present in fecal samples of 90 clinical subjects (comprising 30 CRC patients, 30 with polyps and 30 controls). Our findings revealed a decrease in taxonomic richness among individuals with polyps and CRC, with significant dissimilarities observed among the study groups. We identified significant alterations in the abundance of specific taxa associated with polyps (Streptococcaceae, Lachnoclostridium, and Ralstonia) and CRC (Lactobacillales, Clostridiaceae, Desulfovibrio, SFB, Ruminococcus, and Faecalibacterium). Clostridiaceae exhibited significantly lower abundance in the early stages of CRC. Additionally, our study revealed a positive co-occurrence among underrepresented genera in CRC, while demonstrating a negative co-occurrence between Faecalibacterium and Desulfovibrio, suggesting potential antagonistic relationships. Moreover, we observed variations in taxonomic richness and/or abundance within the polyp and CRC bacteriome linked to polyp size, tumor stage, dyslipidemia, diabetes with metformin use, sex, age, and family history of CRC. These findings provide potential new biomarkers to enhance early CRC diagnosis while also demonstrating how intrinsic host factors contribute to establishing a heterogeneous microbiome in patients with CRC and polyps.

Mucosal Microbiome in Patients with Early Bowel Polyps: Inferences from Short-Read and Long-Read 16S rRNA Sequencing

Numerous studies have correlated dysbiosis in stool microbiota with colorectal cancer (CRC); however, fewer studies have investigated the mucosal microbiome in pre-cancerous bowel polyps. The short-read sequencing of variable regions in the 16S rRNA gene has commonly been used to infer bacterial taxonomy, and this has led, in part, to inconsistent findings between studies. Here, we examined mucosal microbiota from patients who presented with one or more polyps, compared to patients with no polyps, at the time of colonoscopy. We evaluated the results obtained using both short-read and PacBio long-read 16S rRNA sequencing. Neither sequencing technology identified significant differences in microbial diversity measures between patients with or without bowel polyps. Differential abundance measures showed that amplicon sequence variants (ASVs) associated with Ruminococcus gnavus and Escherichia coli were elevated in mucosa from polyp patients, while ASVs associated with Parabacteroides merdae, Veillonella nakazawae, and Sutterella wadsworthensis were relatively decreased. Only R. gnavus was consistently identified using both sequencing technologies as being altered between patients with polyps compared to patients without polyps, suggesting differences in technologies and bioinformatics processing impact study findings. Several of the differentially abundant bacteria identified using either sequencing technology are associated with inflammatory bowel diseases despite these patients being excluded from the current study, which suggests that early bowel neoplasia may be associated with a local inflammatory niche.

Machine learning-based approaches for cancer prediction using microbiome data

Emerging evidence of the relationship between the microbiome composition and the development of numerous diseases, including cancer, has led to an increasing interest in the study of the human microbiome. Technological breakthroughs regarding DNA sequencing methods propelled microbiome studies with a large number of samples, which called for the necessity of more sophisticated data-analytical tools to analyze this complex relationship. The aim of this work was to develop a machine learning-based approach to distinguish the type of cancer based on the analysis of the tissue-specific microbial information, assessing the human microbiome as valuable predictive information for cancer identification. For this purpose, Random Forest algorithms were trained for the classification of five types of cancer-head and neck, esophageal, stomach, colon, and rectum cancers-with samples provided by The Cancer Microbiome Atlas database. One versus all and multi-class classification studies were conducted to evaluate the discriminative capability of the microbial data across increasing levels of cancer site specificity, with results showing a progressive rise in difficulty for accurate sample classification. Random Forest models achieved promising performances when predicting head and neck, stomach, and colon cancer cases, with the latter returning accuracy scores above 90% across the different studies conducted. However, there was also an increased difficulty when discriminating esophageal and rectum cancers, failing to differentiate with adequate results rectum from colon cancer cases, and esophageal from head and neck and stomach cancers. These results point to the fact that anatomically adjacent cancers can be more complex to identify due to microbial similarities. Despite the limitations, microbiome data analysis using machine learning may advance novel strategies to improve cancer detection and prevention, and decrease disease burden.

Fecal Microbiome Associated with Both Colon Adenomas and Lifetime Colorectal Cancer Risk

BACKGROUND

Increasing data indicates the gut flora including bacteria and fungi combined with environmental factors are important in the pathogenesis of colorectal cancer (CRC). Understanding differences in the microbiome in patients with colon neoplasia will foster the development of biomarkers for early detection.

AIMS

Determine the association of microbiome with presence of adenomas and predicted CRC risk.

METHODS

In subjects referred for colonoscopy, the NCI CRC risk assessment tool was completed and stool for microbiome analysis as well as fecal immunochemical test (FIT) were collected. We calculated the microbiome alpha diversity using the Shannon index as well as individual bacterial and fungal species.

RESULTS

Among 34 patients, we identified 10 with one or more adenomas. Only 2 patients were FIT positive. The median predicted lifetime CRC risk was 2.75% and the prevalence of adenoma was higher in the fourth quartile (P < 0.001). The measured alpha diversity was somewhat higher in patients with adenomas (P = 0.07). We identified 4 bacterial species with an increased relative abundance among patients with adenomas [P < 0.5]. Lifetime CRC risk was associated with 2 specific bacterial species, P. distasonis & E. hermannii [P = 0.05 & 0.09, respectively]. No associations were seen with fungal species and adenoma prevalence or lifetime CRC risk.

CONCLUSIONS

In addition to a strong correlation of predicted CRC risk and adenoma prevalence, we also found important differences in specific bacterial species and both adenoma prevalence and CRC risk. Larger trials are needed to potentially implement further data in the clinical setting.

Dysbiosis of a microbiota-immune metasystem in critical illness is associated with nosocomial infections

Critically ill patients in intensive care units experience profound alterations of their gut microbiota that have been linked to a high risk of hospital-acquired (nosocomial) infections and adverse outcomes through unclear mechanisms. Abundant mouse and limited human data suggest that the gut microbiota can contribute to maintenance of systemic immune homeostasis, and that intestinal dysbiosis may lead to defects in immune defense against infections. Here we use integrated systems-level analyses of fecal microbiota dynamics in rectal swabs and single-cell profiling of systemic immune and inflammatory responses in a prospective longitudinal cohort study of critically ill patients to show that the gut microbiota and systemic immunity function as an integrated metasystem, where intestinal dysbiosis is coupled to impaired host defense and increased frequency of nosocomial infections. Longitudinal microbiota analysis by 16s rRNA gene sequencing of rectal swabs and single-cell profiling of blood using mass cytometry revealed that microbiota and immune dynamics during acute critical illness were highly interconnected and dominated by Enterobacteriaceae enrichment, dysregulated myeloid cell responses and amplified systemic inflammation, with a lesser impact on adaptive mechanisms of host defense. Intestinal Enterobacteriaceae enrichment was coupled with impaired innate antimicrobial effector responses, including hypofunctional and immature neutrophils and was associated with an increased risk of infections by various bacterial and fungal pathogens. Collectively, our findings suggest that dysbiosis of an interconnected metasystem between the gut microbiota and systemic immune response may drive impaired host defense and susceptibility to nosocomial infections in critical illness.

Microbiome Profiling from Fecal Immunochemical Test Reveals Microbial Signatures with Potential for Colorectal Cancer Screening

Colorectal cancer (CRC) is the third most common cancer and the second leading cause of cancer deaths worldwide. Early diagnosis of CRC, which saves lives and enables better outcomes, is generally implemented through a two-step population screening approach based on the use of Fecal Immunochemical Test (FIT) followed by colonoscopy if the test is positive. However, the FIT step has a high false positive rate, and there is a need for new predictive biomarkers to better prioritize cases for colonoscopy. Here we used 16S rRNA metabarcoding from FIT positive samples to uncover microbial taxa, taxon co-occurrence and metabolic features significantly associated with different colonoscopy outcomes, underscoring a predictive potential and revealing changes along the path from healthy tissue to carcinoma. Finally, we used machine learning to develop a two-phase classifier which reduces the current false positive rate while maximizing the inclusion of CRC and clinically relevant samples.

Microbiome analysis reveals universal diagnostic biomarkers for colorectal cancer across populations and technologies

The gut microbial dysbiosis is a risk of colorectal cancer (CRC) and some bacteria have been reported as potential markers for CRC diagnosis. However, heterogeneity among studies with different populations and technologies lead to inconsistent results. Here, we investigated six metagenomic profiles of stool samples from healthy controls (HC), colorectal adenoma (CA) and CRC, and six and four genera were consistently altered between CRC and HC or CA across populations, respectively. In FengQ cohort, which composed with 61 HC, 47 CA, and 46 CRC samples, a random forest (RF) model composed of the six genera, denoted as signature-HC, distinguished CRC from HC with an area under the curve (AUC) of 0.84. Similarly, another RF model composed of the four universal genera, denoted as signature-CA, discriminated CRC from CA with an AUC of 0.73. These signatures were further validated in five metagenomic sequencing cohorts and six independent 16S rRNA gene sequencing cohorts. Interestingly, three genera overlapped in the two models (Porphyromonas, Parvimonas and Peptostreptococcus) were with very low abundance in HC and CA, but sharply increased in CRC. A concise RF model on the three genera distinguished CRC from HC or CA with AUC of 0.87 and 0.67, respectively. Functional gene family analysis revealed that Kyoto Encyclopedia of Genes and Genomes Orthogroups categories which were significantly correlated with markers in signature-HC and signature-CA were mapped into pathways related to lipopolysaccharide and sulfur metabolism, which might be vital risk factors of CRC development. Conclusively, our study identified universal bacterial markers across populations and technologies as potential aids in non-invasive diagnosis of CRC.

The crosstalk of the human microbiome in breast and colon cancer: A metabolomics analysis

The human microbiome's role in colon and breast cancer is described in this review. Understanding how the human microbiome and metabolomics interact with breast and colon cancer is the chief area of this study. First, the role of the gut and distal microbiome in breast and colon cancer is investigated, and the direct relationship between microbial dysbiosis and breast and colon cancer is highlighted. This work also focuses on the many metabolomic techniques used to locate prospective biomarkers, make an accurate diagnosis, and research new therapeutic targets for cancer treatment. This review clarifies the influence of anti-tumor medications on the microbiota and the proactive measures that can be taken to treat cancer using a variety of therapies, including radiotherapy, chemotherapy, next-generation biotherapeutics, gene-based therapy, integrated omics technology, and machine learning.

SAMBA: A Multicolor Digital Melting PCR Platform for Rapid Microbiome Profiling

During the past decade, breakthroughs in sequencing technology have provided the basis for studies of the myriad ways in which microbial communities in and on the human body influence human health and disease. In almost every medical specialty, there is now a growing interest in accurate and quantitative profiling of the microbiota for use in diagnostic and therapeutic applications. However, the current next-generation sequencing approach for microbiome profiling is costly, requires laborious library preparation, and is challenging to scale up for routine diagnostics. Split, Amplify, and Melt analysis of BActeria-community (SAMBA), a novel multicolor digital melting polymerase chain reaction platform with unprecedented multiplexing capability is presented, and the capability to distinguish and quantify 16 bacteria species in mixtures is demonstrated. Subsequently, SAMBA is applied to measure the compositions of bacteria in the gut microbiome to identify microbial dysbiosis related to colorectal cancer. This rapid, low cost, and high-throughput approach will enable the implementation of microbiome diagnostics in clinical laboratories and routine medical practice.

Cancer's second genome: Microbial cancer diagnostics and redefining clonal evolution as a multispecies process: Humans and their tumors are not aseptic, and the multispecies nature of cancer modulates clinical care and clonal evolution: Humans and their tumors are not aseptic, and the multispecies nature of cancer modulates clinical care and clonal evolution

The presence and role of microbes in human cancers has come full circle in the last century. Tumors are no longer considered aseptic, but implications for cancer biology and oncology remain underappreciated. Opportunities to identify and build translational diagnostics, prognostics, and therapeutics that exploit cancer's second genome-the metagenome-are manifold, but require careful consideration of microbial experimental idiosyncrasies that are distinct from host-centric methods. Furthermore, the discoveries of intracellular and intra-metastatic cancer bacteria necessitate fundamental changes in describing clonal evolution and selection, reflecting bidirectional interactions with non-human residents. Reconsidering cancer clonality as a multispecies process similarly holds key implications for understanding metastasis and prognosing therapeutic resistance while providing rational guidance for the next generation of bacterial cancer therapies. Guided by these new findings and challenges, this Review describes opportunities to exploit cancer's metagenome in oncology and proposes an evolutionary framework as a first step towards modeling multispecies cancer clonality. Also see the video abstract here: https://youtu.be/-WDtIRJYZSs.

Novel microbiome signatures for non-invasive diagnosis of adenoma recurrence after colonoscopic polypectomy

BACKGROUND

We previously reported a panel of novel faecal microbiome gene markers for diagnosis of colorectal adenoma and cancer.

AIM

To evaluate whether these markers are useful in detecting adenoma recurrence after polypectomy.

METHODS

Subjects were enrolled in a polyp surveillance study from 2009 to 2019. Stool samples were collected before bowel preparation of index colonoscopy (baseline) and surveillance colonoscopy (follow-up). Fusobacterium nucleatum (Fn), Lachnoclostridium marker (m3), Clostridium hathewayi (Ch) and Bacteroides clarus were quantified in baseline and follow-up samples by quantitative polymerase chain reaction (qPCR) to correlate with adenoma recurrence. Recurrence was defined as new adenomas detected >6 months after polypectomy. Faecal immunochemical test (FIT) was performed for comparison.

RESULTS

A total of 161 baseline and 104 follow-up samples were analysed. Among patients with adenoma recurrence, Fn and m3 increased (both P < 0.05) while Ch were unchanged in follow-up versus baseline samples. Among patients without recurrence, Fn and m3 were unchanged while Ch decreased (P < 0.05) in follow-up versus baseline samples. Logistic regression that included changes of m3, Fn and Ch at follow-up compared with baseline achieved an area under receiver operating characteristic curve (AUROC) of 0.95 (95%CI: 0.84-0.99) with 90.0% sensitivity and 87.0% specificity for detecting recurrent adenoma. Combination of m3, Fn and Ch at follow-up sample achieved AUROC of 0.74 (95%CI: 0.65-0.82) with 81.3% sensitivity and 55.4% specificity for detecting recurrent adenoma. FIT showed limited sensitivity (8.3%) in detecting recurrent adenomas.

CONCLUSION

Our combinations of faecal microbiome gene markers can be potentially useful non-invasive tools for detecting adenoma recurrence.

Comparing Published Gut Microbiome Taxonomic Data Across Multinational Studies

BACKGROUND

Nurse researchers are well poised to study the connection of the microbiome to health and disease. Evaluating published microbiome results can assist with study design and hypothesis generation.

OBJECTIVES

This article aims to present and define important analysis considerations in microbiome study planning and to identify genera shared across studies despite methodological differences. This methods article will highlight a workflow that the nurse scientist can use to combine and evaluate taxonomy tables for microbiome study or research proposal planning.

METHODS

We compiled taxonomy tables from 13 published gut microbiome studies that had used Ion Torrent sequencing technology. We searched for studies that had amplified multiple hypervariable (V) regions of the 16S rRNA gene when sequencing the bacteria from healthy gut samples.

RESULTS

We obtained 15 taxonomy tables from the 13 studies, comprised of samples from four continents and eight V regions. Methodology among studies was highly variable, including differences in V regions amplified, geographic location, and population demographics. Nevertheless, of the 354 total genera identified from the 15 data sets, 25 were shared in all V regions and the four continents. When relative abundance differences across the V regions were compared, Dorea and Roseburia were statistically different. Taxonomy tables from Asian subjects had increased average abundances of Prevotella and lowered abundances of Bacteroides compared with the European, North American, and South American study subjects.

DISCUSSION

Evaluating taxonomy tables from previously published literature is essential for study planning. The genera found from different V regions and continents highlight geography and V region as important variables to consider in microbiome study design. The 25 shared genera across the various studies may represent genera commonly found in healthy gut microbiomes. Understanding the factors that may affect the results from a variety of microbiome studies will allow nurse scientists to plan research proposals in an informed manner. This work presents a valuable framework for future cross-study comparisons conducted across the globe.

Artificial intelligence reveals roles of gut microbiota in driving human colorectal cancer evolution

With the rapid development of high-throughput sequencing and artificial intelligence (AI) techniques, gut mucosal microbiota begins to be recognized as critical drivers of human colorectal cancer (CRC). Various AI approaches have been designed to obtain effective information from enormous numbers of microbial cells residing in gut mucosal as well as cancer cells. These mainly include detection of microbial markers for early clinical diagnosis of stage-specific CRC, characterization of pathogenic bacterial activities via genomic and transcriptomic analyses, and prediction of interplay between bacterial drivers and host immune systems. Here I review the current progresses of AI applications in profiling gut microbiomes linked to CRC initiation and development. I further look forward to future AI research for improving our understanding of the roles of gut microbiota in CRC evolution.

Impact of Antibiotic Resistance Genes in Gut Microbiome of Patients With Cirrhosis

BACKGROUND AND AIMS

Cirrhosis is associated with changes in intestinal microbiota that can lead to hepatic encephalopathy (HE) and infections, especially with antibiotic-resistant organisms. However, the impact of gut microbial antibiotic resistance gene (ARG) burden on clinical outcomes is unclear. The aims of the study were to determine the impact of ARGs in cirrhosis-related gut metagenome on outcomes and disease progression, study the effect of rifaximin on ARG burden, and compare ARGs in cirrhosis with chronic kidney disease (CKD) and diabetes.

METHODS

In outpatients with cirrhosis who underwent metagenomics, we evaluated change in ARG abundances with progression and their multivariable impact on 90-day hospitalizations and deaths over 1 year. We also studied ARGs pre- and 8 weeks post-rifaximin in patients with compensated cirrhosis in an open-label trial. Finally, ARGs from CKD and diabetes studies were compared with cirrhosis on machine learning.

RESULTS

A total of 163 patients with cirrhosis (43 compensated, 20 ascites-only, 30 HE-only, 70 both) and 40 controls were included. ARG abundances were higher in cirrhosis versus controls and worsened with advancing cirrhosis severity; 44 patients were hospitalized and 14 died. ARG abundances were associated with hospitalizations and mortality while controlling for cirrhosis complications, medications, and demographics. Rifaximin trial: ARG abundance patterns were minimally affected in 19 patients post-rifaximin. CKD/diabetes comparison: ARG abundance patterns in cirrhosis are distinguishable on machine learning and include more gram-positive ARGs.

CONCLUSIONS

Cirrhosis is associated with high gut microbial ARG gene burden compared with controls, which worsens with disease progression and may be different from CKD and diabetes. ARGs are not affected by rifaximin and are associated with hospitalizations and death.

A Pilot Study Investigating the Expression Levels of Pluripotency-Associated Genes in Rectal Swab Samples for Colorectal Polyp and Cancer Diagnosis and Prognosis

Change in gene expression is inevitable in cancer development. With more studies demonstrating the contributions of cancer stem cells (CSCs) in colorectal cancer (CRC) development, this study is aimed at investigating whether rectal swab specimen serves as a tool for detection of dysregulation of CSC or stem cell (SC) markers and at evaluating its potential as a new promising screening method for high-risk patients. Expression levels of 15 pluripotency-associated genes were assessed by quantitative PCR in 53 rectal swab specimens referred for endoscopic screening. Dysregulated genes and joint panels based on such genes were examined for their diagnostic potentials for both polyp and CRC. Out of 15 genes, Oct4, CD26, c-MYC, and CXCR4 showed significantly differential expression among normal, polyp, and CRC patients. A panel of Oct4 and CD26 showed an AUC value of 0.80 (p = 0.003) in identifying CRC patients from polyp/normal subjects, with sensitivity and specificity of 84.6% and 69.2%. A panel of c-MYC and CXCR4 achieved CRC/polyp identification with an AUC value of 0.79 (p = 0.002), with a sensitivity of 82.8% and specificity of 80.0%. The sensitivity for polyp and CRC was 80.0% and 85.7%, respectively. Further analysis showed that higher c-MYC and CXCR4 level was detected in normal subjects who developed polyps after 5-6 years, in comparison with subjects with no lesion developed, and the AUC of the c-MYC and CXCR4 panel increased to 0.88 (p < 0.001), with sensitivity and specificity of 84.4% and 92.3%, respectively, when these patients were included in the polyp group. This study suggests that the Oct4 and CD26 panel is a promising biomarker for distinguishing CRC from normal and polyp patients, whereas the c-MYC and CXCR4 panel may identify polyp and CRC from normal individuals.

Genomic, Microbial and Immunological Microenvironment of Colorectal Polyps

Simple Summary

Colorectal cancers (CRC) initiate from small cell clusters known as polyps. Colonoscopic surveillance and removal of polyps is an important strategy to prevent CRC progression. Recent advances in sequencing technologies have highlighted genetic mutations in polyps that potentially contribute to CRC development. However, CRC might be considered more than a genetic disease, as emerging evidence describes early changes to immune surveillance and gut microbiota in people with polyps. Here, we review the molecular landscape of colorectal polyps, considering their genomic, microbial and immunological features, and discuss the potential clinical utility of these data.

Abstract

Colorectal cancer (CRC) develops from pre-cancerous cellular lesions in the gut epithelium, known as polyps. Polyps themselves arise through the accumulation of mutations that disrupt the function of key tumour suppressor genes, activate proto-oncogenes and allow proliferation in an environment where immune control has been compromised. Consequently, colonoscopic surveillance and polypectomy are central pillars of cancer control strategies. Recent advances in genomic sequencing technologies have enhanced our knowledge of key driver mutations in polyp lesions that likely contribute to CRC. In accordance with the prognostic significance of Immunoscores for CRC survival, there is also a likely role for early immunological changes in polyps, including an increase in regulatory T cells and a decrease in mature dendritic cell numbers. Gut microbiotas are under increasing research interest for their potential contribution to CRC evolution, and changes in the gut microbiome have been reported from analyses of adenomas. Given that early changes to molecular components of bowel polyps may have a direct impact on cancer development and/or act as indicators of early disease, we review the molecular landscape of colorectal polyps, with an emphasis on immunological and microbial alterations occurring in the gut and propose the potential clinical utility of these data.

Mediterranean Diet to Prevent the Development of Colon Diseases: A Meta-Analysis of Gut Microbiota Studies

Gut microbiota dysbiosis is a common feature in colorectal cancer (CRC) and inflammatory bowel diseases (IBD). Adoption of the Mediterranean diet (MD) has been proposed as a therapeutic approach for the prevention of multiple diseases, and one of its mechanisms of action is the modulation of the microbiota. We aimed to determine whether MD can be used as a preventive measure against cancer and inflammation-related diseases of the gut, based on its capacity to modulate the local microbiota. A joint meta-analysis of publicly available 16S data derived from subjects following MD or other diets and from patients with CRC, IBD, or other gut-related diseases was conducted. We observed that the microbiota associated with MD was enriched in bacteria that promote an anti-inflammatory environment but low in taxa with pro-inflammatory properties capable of altering intestinal barrier functions. We found an opposite trend in patients with intestinal diseases, including cancer. Some of these differences were maintained even when MD was compared to healthy controls without a defined diet. Our findings highlight the unique effects of MD on the gut microbiota and suggest that integrating MD principles into a person’s lifestyle may serve as a preventive method against cancer and other gut-related diseases.

Status quo and future prospects of artificial neural network from the perspective of gastroenterologists

Artificial neural networks (ANNs) are one of the primary types of artificial intelligence and have been rapidly developed and used in many fields. In recent years, there has been a sharp increase in research concerning ANNs in gastrointestinal (GI) diseases. This state-of-the-art technique exhibits excellent performance in diagnosis, prognostic prediction, and treatment. Competitions between ANNs and GI experts suggest that efficiency and accuracy might be compatible in virtue of technique advancements. However, the shortcomings of ANNs are not negligible and may induce alterations in many aspects of medical practice. In this review, we introduce basic knowledge about ANNs and summarize the current achievements of ANNs in GI diseases from the perspective of gastroenterologists. Existing limitations and future directions are also proposed to optimize ANN’s clinical potential. In consideration of barriers to interdisciplinary knowledge, sophisticated concepts are discussed using plain words and metaphors to make this review more easily understood by medical practitioners and the general public.

Identification of microbial markers across populations in early detection of colorectal cancer

Associations between gut microbiota and colorectal cancer (CRC) have been widely investigated. However, the replicable markers for early-stage adenoma diagnosis across multiple populations remain elusive. Here, we perform an integrated analysis on 1056 public fecal samples, to identify adenoma-associated microbial markers for early detection of CRC. After adjusting for potential confounders, Random Forest classifiers are constructed with 11 markers to discriminate adenoma from control (area under the ROC curve (AUC) = 0.80), and 26 markers to discriminate adenoma from CRC (AUC = 0.89), respectively. Moreover, we validate the classifiers in two independent cohorts achieving AUCs of 0.78 and 0.84, respectively. Functional analysis reveals that the altered microbiome is characterized with increased ADP-L-glycero-beta-D-manno-heptose biosynthesis in adenoma and elevated menaquinone-10 biosynthesis in CRC. These findings are validated in a newly-collected cohort of 43 samples using quantitative real-time PCR. This work proves the validity of adenoma-specific markers across multi-populations, which would contribute to the early diagnosis and treatment of CRC.

Gut microbiota and artificial intelligence approaches: A scoping review

This article aims to provide a thorough overview of the use of Artificial Intelligence (AI) techniques in studying the gut microbiota and its role in the diagnosis and treatment of some important diseases. The association between microbiota and diseases, together with its clinical relevance, is still difficult to interpret. The advances in AI techniques, such as Machine Learning (ML) and Deep Learning (DL), can help clinicians in processing and interpreting these massive data sets. Two research groups have been involved in this Scoping Review, working in two different areas of Europe: Florence and Sarajevo. The papers included in the review describe the use of ML or DL methods applied to the study of human gut microbiota. In total, 1109 papers were considered in this study. After elimination, a final set of 16 articles was considered in the scoping review. Different AI techniques were applied in the reviewed papers. Some papers applied ML, while others applied DL techniques. 11 papers evaluated just different ML algorithms (ranging from one to eight algorithms applied to one dataset). The remaining five papers examined both ML and DL algorithms. The most applied ML algorithm was Random Forest and it also exhibited the best performances.

BiomMiner: An advanced exploratory microbiome analysis and visualization pipeline

Current microbiome applications require substantial bioinformatics expertise to execute. As microbiome clinical diagnostics are being developed, there is a critical need to implement computational tools and applications that are user-friendly for the medical community to understand microbiome correlation with the health. To address this need, we have developed BiomMiner (pronounced as “biominer”), an automated pipeline that provides a comprehensive analysis of microbiome data. The pipeline finds taxonomic signatures of microbiome data and compiles actionable clinical report that allows clinicians and biomedical scientists to efficiently perform statistical analysis and data mining on the large microbiome datasets. BiomMiner generates web-enabled visualization of the analysis results and is specifically designed to facilitate the use of microbiome datasets in clinical applications.

A Framework for Effective Application of Machine Learning to Microbiome-Based Classification Problems

Machine learning (ML) modeling of the human microbiome has the potential to identify microbial biomarkers and aid in the diagnosis of many diseases such as inflammatory bowel disease, diabetes, and colorectal cancer. Progress has been made toward developing ML models that predict health outcomes using bacterial abundances, but inconsistent adoption of training and evaluation methods call the validity of these models into question. Furthermore, there appears to be a preference by many researchers to favor increased model complexity over interpretability. To overcome these challenges, we trained seven models that used fecal 16S rRNA sequence data to predict the presence of colonic screen relevant neoplasias (SRNs) (n = 490 patients, 261 controls and 229 cases). We developed a reusable open-source pipeline to train, validate, and interpret ML models. To show the effect of model selection, we assessed the predictive performance, interpretability, and training time of L2-regularized logistic regression, L1- and L2-regularized support vector machines (SVM) with linear and radial basis function kernels, a decision tree, random forest, and gradient boosted trees (XGBoost). The random forest model performed best at detecting SRNs with an area under the receiver operating characteristic curve (AUROC) of 0.695 (interquartile range [IQR], 0.651 to 0.739) but was slow to train (83.2 h) and not inherently interpretable. Despite its simplicity, L2-regularized logistic regression followed random forest in predictive performance with an AUROC of 0.680 (IQR, 0.625 to 0.735), trained faster (12 min), and was inherently interpretable. Our analysis highlights the importance of choosing an ML approach based on the goal of the study, as the choice will inform expectations of performance and interpretability.IMPORTANCE Diagnosing diseases using machine learning (ML) is rapidly being adopted in microbiome studies. However, the estimated performance associated with these models is likely overoptimistic. Moreover, there is a trend toward using black box models without a discussion of the difficulty of interpreting such models when trying to identify microbial biomarkers of disease. This work represents a step toward developing more-reproducible ML practices in applying ML to microbiome research. We implement a rigorous pipeline and emphasize the importance of selecting ML models that reflect the goal of the study. These concepts are not particular to the study of human health but can also be applied to environmental microbiology studies.

Oral microbiome-systemic link studies: perspectives on current limitations and future artificial intelligence-based approaches

In the past decade, there has been a tremendous increase in studies on the link between oral microbiome and systemic diseases. However, variations in study design and confounding variables across studies often lead to inconsistent observations. In this narrative review, we have discussed the potential influence of study design and confounding variables on the current sequencing-based oral microbiome-systemic disease link studies. The current limitations of oral microbiome-systemic link studies on type 2 diabetes mellitus, rheumatoid arthritis, pregnancy, atherosclerosis, and pancreatic cancer are discussed in this review, followed by our perspective on how artificial intelligence (AI), particularly machine learning and deep learning approaches, can be employed for predicting systemic disease and host metadata from the oral microbiome. The application of AI for predicting systemic disease as well as host metadata requires the establishment of a global database repository with microbiome sequences and annotated host metadata. However, this task requires collective efforts from researchers working in the field of oral microbiome to establish more comprehensive datasets with appropriate host metadata. Development of AI-based models by incorporating consistent host metadata will allow prediction of systemic diseases with higher accuracies, bringing considerable clinical benefits.

The Bacterial Microbiota of Gastrointestinal Cancers: Role in Cancer Pathogenesis and Therapeutic Perspectives

The microbiota has an essential role in the pathogenesis of many gastrointestinal diseases including cancer. This effect is mediated through different mechanisms such as damaging DNA, activation of oncogenic pathways, production of carcinogenic metabolites, stimulation of chronic inflammation, and inhibition of antitumor immunity. Recently, the concept of "pharmacomicrobiomics" has emerged as a new field concerned with exploring the interplay between drugs and microbes. Mounting evidence indicates that the microbiota and their metabolites have a major impact on the pharmacodynamics and therapeutic responses toward anticancer drugs including conventional chemotherapy and molecular-targeted therapeutics. In addition, microbiota appears as an attractive target for cancer prevention and treatment. In this review, we discuss the role of bacterial microbiota in the pathogenesis of different cancer types affecting the gastrointestinal tract system. We also scrutinize the evidence regarding the role of microbiota in anticancer drug responses. Further, we discuss the use of probiotics, fecal microbiota transplantation, and antibiotics, either alone or in combination with anticancer drugs for prevention and treatment of gastrointestinal tract cancers.

Influence of the Gut Microbiome, Diet, and Environment on Risk of Colorectal Cancer

Researchers have discovered associations between elements of the intestinal microbiome (including specific microbes, signaling pathways, and microbiota-related metabolites) and risk of colorectal cancer (CRC). However, it is unclear whether changes in the intestinal microbiome contribute to the development of sporadic CRC or result from it. Changes in the intestinal microbiome can mediate or modify the effects of environmental factors on risk of CRC. Factors that affect risk of CRC also affect the intestinal microbiome, including overweight and obesity; physical activity; and dietary intake of fiber, whole grains, and red and processed meat. These factors alter microbiome structure and function, along with the metabolic and immune pathways that mediate CRC development. We review epidemiologic and laboratory evidence for the influence of the microbiome, diet, and environmental factors on CRC incidence and outcomes. Based on these data, features of the intestinal microbiome might be used for CRC screening and modified for chemoprevention and treatment. Integrated prospective studies are urgently needed to investigate these strategies.

Current State of and Future Opportunities for Prediction in Microbiome Research: Report from the Mid-Atlantic Microbiome Meet-up in Baltimore on 9 January 2019

Accurate predictions across multiple fields of microbiome research have far-reaching benefits to society, but there are few widely accepted quantitative tools to make accurate predictions about microbial communities and their functions. More discussion is needed about the current state of microbiome analysis and the tools required to overcome the hurdles preventing development and implementation of predictive analyses. We summarize the ideas generated by participants of the Mid-Atlantic Microbiome Meet-up in January 2019. While it was clear from the presentations that most fields have advanced beyond simple associative and descriptive analyses, most fields lack essential elements needed for the development and application of accurate microbiome predictions. Participants stressed the need for standardization, reproducibility, and accessibility of quantitative tools as key to advancing predictions in microbiome analysis. We highlight hurdles that participants identified and propose directions for future efforts that will advance the use of prediction in microbiome research.