Performance of risk prediction for inflammatory bowel disease based on genotyping platform and genomic risk score method

MIP4IBD: An Easy and Rapid Genotyping-by-Sequencing Assay for the Inflammatory Bowel Diseases Risk Loci

BACKGROUND

Inflammatory bowel diseases (IBD) are polygenic, with many genetic variants contributing to disease risk. Knowing the genotype of specific variants or calculating a combined genetic risk score benefits translational and functional research. To address this, we developed MIP4IBD, a flexible and cost-effective genotyping-by-sequencing assay using molecular inversion probes (MIPs).

METHODS

The assay targets 463 IBD risk variants, and 77 additional relevant variants. Molecular inversion probes capture and library preparation were optimized using 15 IBD DNA samples, comparing genotypes with immunochip. A custom GitHub pipeline was created for data processing, performance testing, and genotype calling. The final design was validated on a larger scale (149 IBD patients, 104 non-IBD controls, and 3 external cell lines), incorporating post hoc quality control criteria.

RESULTS

The assay achieved a 3.5-day turnaround time at €15 per sample with optimal sample throughput, demonstrating a 92.6% success rate in variant capture and genotype concordance rates of 99.3% and 99.6% with Infinium Global Screening Array24 BeadChip and WGS, respectively. A downstream application involved the calculation of a weighted IBD polygenic risk score (PRS), which was significantly higher in IBD patients than controls (mean 0.42 vs -0.49, P = 1.95E-11). Individuals in the highest PRS quartile had a 15.7-fold (95% CI: 6.5-38.3) risk of developing IBD and an earlier age of onset (26 vs 37 years, P = 0.02), compared to the lowest quartile.

CONCLUSIONS

MIP4IBD is a validated, scalable genotyping assay targeting IBD risk loci, with an integrated bioinformatics pipeline from sequencing data to genotypes and PRS calculation. Its cost-effectiveness and flexibility for additional variants make it particularly appealing for translational and clinical applications.

The Contribution of Genetic and Epigenetic Factors: An Emerging Concept in the Assessment and Prognosis of Inflammatory Bowel Diseases

Inflammatory bowel disease (IBD) represents heterogeneous and relapsing intestinal conditions with a severe impact on the quality of life of individuals and a continuously increasing prevalence. In recent years, the development of sequencing technology has provided new means of exploring the complex pathogenesis of IBD. An ideal solution is represented by the approach of precision medicine that investigates multiple cellular and molecular interactions, which are tools that perform a holistic, systematic, and impartial analysis of the genomic, transcriptomic, proteomic, metabolomic, and microbiomics sets. Hence, it has led to the orientation of current research towards the identification of new biomarkers that could be successfully used in the management of IBD patients. Multi-omics explores the dimension of variation in the characteristics of these diseases, offering the advantage of understanding the cellular and molecular mechanisms that affect intestinal homeostasis for a much better prediction of disease development and choice of treatment. This review focuses on the progress made in the field of prognostic and predictive biomarkers, highlighting the limitations, challenges, and also the opportunities associated with the application of genomics and epigenomics technologies in clinical practice.

Advances in Inflammatory Bowel Disease Diagnostics: Machine Learning and Genomic Profiling Reveal Key Biomarkers for Early Detection

This study, utilizing high-throughput technologies and Machine Learning (ML), has identified gene biomarkers and molecular signatures in Inflammatory Bowel Disease (IBD). We could identify significant upregulated or downregulated genes in IBD patients by comparing gene expression levels in colonic specimens from 172 IBD patients and 22 healthy individuals using the GSE75214 microarray dataset. Our ML techniques and feature selection methods revealed six Differentially Expressed Gene (DEG) biomarkers (VWF, IL1RL1, DENND2B, MMP14, NAAA, and PANK1) with strong diagnostic potential for IBD. The Random Forest (RF) model demonstrated exceptional performance, with accuracy, F1-score, and AUC values exceeding 0.98. Our findings were rigorously validated with independent datasets (GSE36807 and GSE10616), further bolstering their credibility and showing favorable performance metrics (accuracy: 0.841, F1-score: 0.734, AUC: 0.887). Our functional annotation and pathway enrichment analysis provided insights into crucial pathways associated with these dysregulated genes. DENND2B and PANK1 were identified as novel IBD biomarkers, advancing our understanding of the disease. The validation in independent cohorts enhances the reliability of these findings and underscores their potential for early detection and personalized treatment of IBD. Further exploration of these genes is necessary to fully comprehend their roles in IBD pathogenesis and develop improved diagnostic tools and therapies. This study significantly contributes to IBD research with valuable insights, potentially greatly enhancing patient care.

The Genetics of Inflammatory Bowel Disease

The genetic background of inflammatory bowel disease, both Crohn's disease and ulcerative colitis, has been known for more than 2 decades. In the last 20 years, genome-wide association studies have dramatically increased our knowledge on the genetics of inflammatory bowel disease with more than 200 risk genes having been identified. Paralleling this increasing knowledge, the armamentarium of inflammatory bowel disease medications has been growing constantly. With more available therapeutic options, treatment decisions become more complex, with still many patients experiencing a debilitating disease course and a loss of response to treatment over time. With a better understanding of the disease, more effective personalized treatment strategies are looming on the horizon. Genotyping has long been considered a strategy for treatment decisions, such as the detection of thiopurine S-methyltransferase and nudix hydrolase 15 polymorphisms before the initiation of azathioprine. However, although many risk genes have been identified in inflammatory bowel disease, a substantial impact of genetic risk assessment on therapeutic strategies and disease outcome is still missing. In this review, we discuss the genetic background of inflammatory bowel disease, with a particular focus on the latest advances in the field and their potential impact on management decisions.

The promise of explainable deep learning for omics data analysis: Adding new discovery tools to AI

Deep learning has already revolutionised the way a wide range of data is processed in many areas of daily life. The ability to learn abstractions and relationships from heterogeneous data has provided impressively accurate prediction and classification tools to handle increasingly big datasets. This has a significant impact on the growing wealth of omics datasets, with the unprecedented opportunity for a better understanding of the complexity of living organisms. While this revolution is transforming the way these data are analyzed, explainable deep learning is emerging as an additional tool with the potential to change the way biological data is interpreted. Explainability addresses critical issues such as transparency, so important when computational tools are introduced especially in clinical environments. Moreover, it empowers artificial intelligence with the capability to provide new insights into the input data, thus adding an element of discovery to these already powerful resources. In this review, we provide an overview of the transformative effects explainable deep learning is having on multiple sectors, ranging from genome engineering and genomics, from radiomics to drug design and clinical trials. We offer a perspective to life scientists, to better understand the potential of these tools, and a motivation to implement them in their research, by suggesting learning resources they can use to move their first steps in this field.

The Pediatric Crohn Disease Morbidity Index (PCD-MI): Development of a Tool to Assess Long-Term Disease Burden Using a Data-Driven Approach

BACKGROUND/OBJECTIVE

Heterogeneity and chronicity of Crohn disease (CD) make prediction of outcomes difficult. To date, no longitudinal measure can quantify burden over a patient's disease course, preventing assessment and integration into predictive modeling. Here, we aimed to demonstrate the feasibility of constructing a data driven, longitudinal disease burden score.

METHODS

Literature was reviewed for tools used in assessment of CD activity. Themes were identified to construct a pediatric CD morbidity index (PCD-MI). Scores were assigned to variables. Data were extracted automatically from the electronic patient records at Southampton Children's Hospital, diagnosed from 2012 to 2019 (inclusive). PCD-MI scores were calculated, adjusted for duration of follow up and assessed for variation (ANOVA) and distribution (Kolmogorov-Smirnov).

RESULTS

Nineteen clinical/biological features across five themes were included in the PCD-MI including blood/fecal/radiological/endoscopic results, medication usage, surgery, growth parameters, and extraintestinal manifestations. Maximal score was 100 after accounting for follow-up duration. PCD-MI was assessed in 66 patients, mean age 12.5 years. Following quality filtering, 9528 blood/fecal test results and 1309 growth measures were included. Mean PCD-MI score was 14.95 (range 2.2-32.5); data were normally distributed ( P = 0.2) with 25% of patients having a PCD-MI < 10. There was no difference in the mean PCD-MI when split by year of diagnosis, F -statistic 1.625, P = 0.147.

CONCLUSIONS

PCD-MI is a calculatable measure for a cohort of patients diagnosed over an 8-year period, integrating a wide-range of data with potential to determine high or low disease burden. Future iterations of the PCD-MI require refinement of included features, optimized scores, and validation on external cohorts.

New pattern of individualized management of chronic diseases: focusing on inflammatory bowel diseases and looking to the future

Non-infectious chronic diseases, especially inflammatory bowel diseases (IBDs), hypertension, and diabetes mellitus, are characterized by a prolonged and multisystemic course, and their incidence increases annually, usually causing serious economic burden and psychological stress for patients. Therefore, these diseases deserve scientific and consistent disease management. In addition, the lack of a comprehensive "early disease clues tracking-personalized treatment system-follow-up" model in hospitals also exacerbates this dilemma. Based on these facts, we propose an individualized prediction management system for IBDs based on chronic diseases, focusing on the established IBDs-related prediction models and summarizing their advantages and disadvantages. We call on researchers to pay attention to the integration of models with clinical practice and the continuous correction of models to achieve truly individualized medical treatment for chronic diseases, thus providing substantial value for the rapid diagnosis and adequate treatment of chronic diseases such as IBDs, which follow the "relapse-remission" disease model, and realizing long-term drug use and precise disease management for patients. The goal is to achieve a new level of chronic disease management by scientifically improving long-term medication, precise disease management, and individualized medical treatment, effectively prolonging the remission period and reducing morbidity and disability rates.

The genetics of non-monogenic IBD

Inflammatory bowel disease (IBD), with Crohn's disease and ulcerative colitis as main subtypes, is a prototypical multifactorial disease with both genetic and environmental factors involved. Genetically, IBD covers a wide spectrum from monogenic to polygenic forms. In polygenic disease, many genetic variants each contribute a small amount to disease risk. With the advent of genome-wide association studies (GWAS), it became possible to find these variants and corresponding genes, leading so far to the discovery of ca 240 loci associated with IBD. Together, these however explain only 20-25% of the heritability of IBD, leaving a large portion unaccounted for. This missing heritability might be hidden in common variants with even lower effect than the ones currently found through GWAS, but also in rare variants which can be found through large-scale sequencing studies or potentially in multiplex families. In this review, we will give an overview of the current knowledge about the genetics of non-monogenic IBD and how it differs from the monogenic form(s), and future perspectives. The history of IBD genetic studies from twin studies over linkage studies to GWAS, and finally large-scale sequencing studies and the revisiting of multiplex families will be discussed.

Assessing the effect of interaction between gut microbiome and inflammatory bowel disease on the risks of depression

Background

Gut microbiome and inflammatory bowel disease (IBD) are implicated in the development of depression, but the effect of their interactions on the risk of depression remains unclear. We aim to analyze the effect of interactions between gut microbiome and IBD on the risk of depression, and explore candidate genes involving the interactions.

Methods

Using the individual genotype and depression traits data from the UK Biobank, we calculated the polygenetic risk scores (PRS) of 114 gut microbiome, ulcerative colitis (UC), Crohn's disease (CD), and total IBD (CD + UC) respectively. The effects of interactions between gut microbiome and IBD on depression were assessed through a linear regression model. Moreover, for observed significant interactions between gut microbiome PRS and IBD PRS, PLINK software was used to test pair-wise single nucleotide polymorphisms (SNPs) interaction of corresponding gut microbiome PRS and IBD PRS on depression.

Results

We found 64 candidate interactions between gut microbiome and IBD on four phenotypes of depression, such as F_Lachnospiraceae (RNT) × (CD + UC) for patient health questionnaire-9 (PHQ-9) score (P = 1.48 × 10^-3), F_Veillonellaceae (HB) × UC for self-reported depression (P = 2.83 × 10^-3) and P_Firmicutes (RNT) × CD for age at first episode of depression (P = 8.50 × 10^-3). We observed interactions of gut-microbiome-associated SNPs × IBD-associated SNPs, such as G_Alloprevotella (HB)-associated rs147650986 (GPM6A) × IBD-associated rs114471990 (QRICH1) (P = 2.26 × 10^-4).

Conclusion

Our results support the effects of interactions between gut microbiome and IBD on depression risk, and reported several novel candidate genes for depression.

Crohn's disease in endoscopic remission, obesity, and cases of high genetic risk demonstrates overlapping shifts in the colonic mucosal-luminal interface microbiome

BACKGROUND

Crohn's disease (CD) patients demonstrate distinct intestinal microbial compositions and metabolic characteristics compared to unaffected controls. However, the impact of inflammation and underlying genetic risk on these microbial profiles and their relationship to disease phenotype are unclear. We used lavage sampling to characterize the colonic mucosal-luminal interface (MLI) microbiome of CD patients in endoscopic remission and unaffected controls relative to obesity, disease genetics, and phenotype.

METHODS

Cecum and sigmoid colon were sampled from 110 non-CD controls undergoing screening colonoscopy who were stratified by body mass index and 88 CD patients in endoscopic remission (396 total samples). CD polygenic risk score (GRS) was calculated using 186 known CD variants. MLI pellets were analyzed by 16S ribosomal RNA gene sequencing, and supernatants by untargeted liquid chromatography-mass spectrometry.

RESULTS

CD and obesity were each associated with decreased cecal and sigmoid MLI bacterial diversity and distinct bacterial composition compared to controls, including expansion of Escherichia/Shigella. Cecal and sigmoid dysbiosis indices for CD were significantly greater in obese controls than non-overweight controls. CD, but not obesity, was characterized by altered biogeographic relationship between the sigmoid and cecum. GRS was associated with select taxonomic shifts that overlapped with changes seen in CD compared to controls including Fusobacterium enrichment. Stricturing or penetrating Crohn's disease behavior was characterized by lower MLI bacterial diversity and altered composition, including reduced Faecalibacterium, compared to uncomplicated CD. Taxonomic profiles including reduced Parasutterella were associated with clinical disease progression over a mean follow-up of 3.7 years. Random forest classifiers using MLI bacterial abundances could distinguish disease state (area under the curve (AUC) 0.93), stricturing or penetrating Crohn's disease behavior (AUC 0.82), and future clinical disease progression (AUC 0.74). CD patients showed alterations in the MLI metabolome including increased cholate:deoxycholate ratio compared to controls.

CONCLUSIONS

Obesity, CD in endoscopic remission, and high CD genetic risk have overlapping colonic mucosal-luminal interface (MLI) microbiome features, suggesting a shared microbiome contribution to CD and obesity which may be influenced by genetic factors. Microbial profiling during endoscopic remission predicted Crohn's disease behavior and progression, supporting that MLI sampling could offer unique insight into CD pathogenesis and provide novel prognostic biomarkers.

Early Diagnosis, Early Stratification, and Early Intervention to Deliver Precision Medicine in IBD

Despite huge advances in understanding the molecular basis of IBD, clinical management has continued to rely on a "trial and error" approach. In addition, a therapeutic ceiling has emerged whereby even the most effective interventions are only beneficial for approximately 30% of patients. Consequently, several tools have been developed to aid stratification and guide treatment-decisions. We review the potential application for many of these precision medicine approaches, which are now almost within reach. We highlight the importance of early action (and avoiding inaction) to ensure the best outcomes for patients and how combining early action with precision tools will likely ensure the right treatment is delivered at the right time and place for each individual person living with IBD. The lack of clinical impact to date from precision medicine, despite much hype and investment, should be tempered with the knowledge that clinical translation can take a long time, and many promising breakthroughs might be ready for clinical implementation in the near future. We discuss some of the remaining challenges and barriers to overcome for clinical adoption. We also highlight that early recognition, early diagnosis, early stratification, and early intervention go hand in hand with precision medicine tools. It is the combination of these approaches that offer the greatest opportunity to finally deliver on the promise of precision medicine in IBD.

Genes vs environment in inflammatory bowel disease: an update

INTRODUCTION

Inflammatory bowel diseases (IBDs) are known to be caused by a combination of genetic and environmental factors that vary in their influence on the development of the disease. Environmental exposures seem to influence IBD susceptibility, whereas genetic background is thought to modulate the impact of the environment on disease course and phenotype.

AREAS COVERED

A broad review of the involvement of genes and the environment in IBD pathogenesis was performed, and information regarding the main genetic and environmental factors - categorized into lifestyle factors, drugs, diet, and microbes - was updated. Monogenic very early onset IBD (VEO-IBD) was also discussed.

EXPERT OPINION

In the upcoming years, better understanding of gene-environment interactions will contribute to the possibility of a better prediction of disease course, response to therapy, and therapy-related adverse events with the final goal of personalized and more efficient patient management.

Parental inflammatory bowel disease and autism in children

Evidence linking parental inflammatory bowel disease (IBD) with autism in children is inconclusive. We conducted four complementary studies to investigate associations between parental IBD and autism in children, and elucidated their underlying etiology. Conducting a nationwide population-based cohort study using Swedish registers, we found evidence of associations between parental diagnoses of IBD and autism in children. Polygenic risk score analyses of the Avon Longitudinal Study of Parents and Children suggested associations between maternal genetic liability to IBD and autistic traits in children. Two-sample Mendelian randomization analyses provided evidence of a potential causal effect of genetic liability to IBD, especially ulcerative colitis, on autism. Linkage disequilibrium score regression did not indicate a genetic correlation between IBD and autism. Triangulating evidence from these four complementary approaches, we found evidence of a potential causal link between parental, particularly maternal, IBD and autism in children. Perinatal immune dysregulation, micronutrient malabsorption and anemia may be implicated.

A systematic review and functional bioinformatics analysis of genes associated with Crohn's disease identify more than 120 related genes

BACKGROUND

Crohn's disease is one of the two categories of inflammatory bowel diseases that affect the gastrointestinal tract. The heritability estimate has been reported to be 0.75. Several genes linked to Crohn's disease risk have been identified using a plethora of strategies such as linkage-based studies, candidate gene association studies, and lately through genome-wide association studies (GWAS). Nevertheless, to our knowledge, a compendium of all the genes that have been associated with CD is lacking.

METHODS

We conducted functional analyses of a gene set generated from a systematic review where genes potentially related to CD found in the literature were analyzed and classified depending on the genetic evidence reported and putative biological function. For this, we retrieved and analyzed 2496 abstracts comprising 1067 human genes plus 22 publications regarding 133 genes from GWAS Catalog. Then, each gene was curated and categorized according to the type of evidence associated with Crohn's disease.

RESULTS

We identified 126 genes associated with Crohn's disease risk by specific experiments. Additionally, 71 genes were recognized associated through GWAS alone, 18 to treatment response, 41 to disease complications, and 81 to related diseases. Bioinformatic analysis of the 126 genes supports their importance in Crohn's disease and highlights genes associated with specific aspects such as symptoms, drugs, and comorbidities. Importantly, most genes were not included in commercial genetic panels suggesting that Crohn's disease is genetically underdiagnosed.

CONCLUSIONS

We identified a total of 126 genes from PubMed and 71 from GWAS that showed evidence of association to diagnosis, 18 to treatment response, and 41 to disease complications in Crohn's disease. This prioritized gene catalog can be explored at http://victortrevino.bioinformatics.mx/CrohnDisease .

Clinical Phenotypes and Outcomes in Monogenic Versus Non-monogenic Very Early Onset Inflammatory Bowel Disease

BACKGROUND AND AIMS

Over 80 monogenic causes of very early onset inflammatory bowel disease [VEOIBD] have been identified. Prior reports of the natural history of VEOIBD have not considered monogenic disease status. The objective of this study is to describe clinical phenotypes and outcomes in a large single-centre cohort of patients with VEOIBD and universal access to whole exome sequencing [WES].

METHODS

Patients receiving IBD care at a single centre were prospectively enrolled in a longitudinal data repository starting in 2012. WES was offered with enrollment. Enrolled patients were filtered by age of diagnosis <6 years to comprise a VEOIBD cohort. Monogenic disease was identified by filtering proband variants for rare, loss-of-function, or missense variants in known VEOIBD genes inherited according to standard Mendelian inheritance patterns.

RESULTS

This analysis included 216 VEOIBD patients, followed for a median of 5.8 years. Seventeen patients [7.9%] had monogenic disease. Patients with monogenic IBD were younger at diagnosis and were more likely to have Crohn's disease phenotype with higher rates of stricturing and penetrating disease and extraintestinal manifestations. Patients with monogenic disease were also more likely to experience outcomes of intensive care unit [ICU] hospitalisation, gastrostomy tube, total parenteral nutrition use, stunting at 3-year follow-up, haematopoietic stem cell transplant, and death. A total of 41 patients [19.0%] had infantile-onset disease. After controlling for monogenic disease, patients with infantile-onset IBD did not have increased risk for most severity outcomes.

CONCLUSIONS

Monogenic disease is an important driver of disease severity in VEOIBD. WES is a valuable tool in prognostication and management of VEOIBD.

Local genetic variation of inflammatory bowel disease in Basque population and its effect in risk prediction

Inflammatory bowel disease (IBD) is characterised by chronic inflammation of the gastrointestinal tract. Although its aetiology remains unknown, environmental and genetic factors are involved in its development. Regarding genetics, more than 200 loci have been associated with IBD but the transferability of those signals to the Basque population living in Northern Spain, a population with distinctive genetic background, remains unknown. We have analysed 5,411,568 SNPs in 498 IBD cases and 935 controls from the Basque population. We found 33 suggestive loci (p < 5 × 10⁻⁶) in IBD and its subtypes, namely Crohn’s Disease (CD) and Ulcerative Colitis (UC), detecting a genome-wide significant locus located in HLA region in patients with UC. Those loci contain previously associated genes with IBD (IL23R, JAK2 or HLA genes) and new genes that could be involved in its development (AGT, BZW2 or FSTL1). The overall genetic correlation between European populations and Basque population was high in IBD and CD, while in UC was lower. Finally, the use of genetic risk scores based on previous GWAS findings reached area under the curves > 0.68. In conclusion, we report on the genetic architecture of IBD in the Basque population, and explore the performance of European-descent genetic risk scores in this population.

Machine Learning Modeling from Omics Data as Prospective Tool for Improvement of Inflammatory Bowel Disease Diagnosis and Clinical Classifications

Research of inflammatory bowel disease (IBD) has identified numerous molecular players involved in the disease development. Even so, the understanding of IBD is incomplete, while disease treatment is still far from the precision medicine. Reliable diagnostic and prognostic biomarkers in IBD are limited which may reduce efficient therapeutic outcomes. High-throughput technologies and artificial intelligence emerged as powerful tools in search of unrevealed molecular patterns that could give important insights into IBD pathogenesis and help to address unmet clinical needs. Machine learning, a subtype of artificial intelligence, uses complex mathematical algorithms to learn from existing data in order to predict future outcomes. The scientific community has been increasingly employing machine learning for the prediction of IBD outcomes from comprehensive patient data-clinical records, genomic, transcriptomic, proteomic, metagenomic, and other IBD relevant omics data. This review aims to present fundamental principles behind machine learning modeling and its current application in IBD research with the focus on studies that explored genomic and transcriptomic data. We described different strategies used for dealing with omics data and outlined the best-performing methods. Before being translated into clinical settings, the developed machine learning models should be tested in independent prospective studies as well as randomized controlled trials.

DeepCOMBI: explainable artificial intelligence for the analysis and discovery in genome-wide association studies

Deep learning has revolutionized data science in many fields by greatly improving prediction performances in comparison to conventional approaches. Recently, explainable artificial intelligence has emerged as an area of research that goes beyond pure prediction improvement by extracting knowledge from deep learning methodologies through the interpretation of their results. We investigate such explanations to explore the genetic architectures of phenotypes in genome-wide association studies. Instead of testing each position in the genome individually, the novel three-step algorithm, called DeepCOMBI, first trains a neural network for the classification of subjects into their respective phenotypes. Second, it explains the classifiers’ decisions by applying layer-wise relevance propagation as one example from the pool of explanation techniques. The resulting importance scores are eventually used to determine a subset of the most relevant locations for multiple hypothesis testing in the third step. The performance of DeepCOMBI in terms of power and precision is investigated on generated datasets and a 2007 study. Verification of the latter is achieved by validating all findings with independent studies published up until 2020. DeepCOMBI is shown to outperform ordinary raw P-value thresholding and other baseline methods. Two novel disease associations (rs10889923 for hypertension, rs4769283 for type 1 diabetes) were identified.

Prognostic biomarkers to identify patients likely to develop severe Crohn's disease: a systematic review

BACKGROUND

Identification of biomarkers that predict severe Crohn's disease is an urgent unmet research need, but existing research is piecemeal and haphazard.

OBJECTIVE

To identify biomarkers that are potentially able to predict the development of subsequent severe Crohn's disease.

DESIGN

This was a prognostic systematic review with meta-analysis reserved for those potential predictors with sufficient existing research (defined as five or more primary studies).

DATA SOURCES

PubMed and EMBASE searched from inception to 1 January 2016, updated to 1 January 2018.

REVIEW METHODS

Eligible studies were studies that compared biomarkers in patients who did or did not subsequently develop severe Crohn's disease. We excluded biomarkers that had insufficient research evidence. A clinician and two statisticians independently extracted data relating to predictors, severe disease definitions, event numbers and outcomes, including odds/hazard ratios. We assessed risk of bias. We searched for associations with subsequent severe disease rather than precise estimates of strength. A random-effects meta-analysis was performed separately for odds ratios.

RESULTS

In total, 29,950 abstracts yielded just 71 individual studies, reporting 56 non-overlapping cohorts. Five clinical biomarkers (Montreal behaviour, age, disease duration, disease location and smoking), two serological biomarkers (anti-Saccharomyces cerevisiae antibodies and anti-flagellin antibodies) and one genetic biomarker (nucleotide-binding oligomerisation domain-containing protein 2) displayed statistically significant prognostic potential. Overall, the strongest association with subsequent severe disease was identified for Montreal B2 and B3 categories (odds ratio 4.09 and 6.25, respectively).

LIMITATIONS

Definitions of severe disease varied widely, and some studies confounded diagnosis and prognosis. Risk of bias was rated as 'high' in 92% of studies overall. Some biomarkers that are used regularly in daily practice, for example C-reactive protein, were studied too infrequently for meta-analysis.

CONCLUSIONS

Research for individual biomarkers to predict severe Crohn's disease is scant, heterogeneous and at a high risk of bias. Despite a large amount of potential research, we encountered relatively few biomarkers with data sufficient for meta-analysis, identifying only eight biomarkers with potential predictive capability.

FUTURE WORK

We will use existing data sets to develop and then validate a predictive model based on the potential predictors identified by this systematic review. Contingent on the outcome of that research, a prospective external validation may prove clinically desirable.

STUDY REGISTRATION

This study is registered as PROSPERO CRD42016029363.

FUNDING

This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 25, No. 45. See the NIHR Journals Library website for further project information.

Current Developments of Clinical Sequencing and the Clinical Utility of Polygenic Risk Scores in Inflammatory Diseases

In this mini-review, we highlight selected research by the Deutsche Forschungsgemeinschaft (DFG) Cluster of Excellence “Precision Medicine in Chronic Inflammation” focusing on clinical sequencing and the clinical utility of polygenic risk scores as well as its implication on precision medicine in the field of the inflammatory diseases inflammatory bowel disease, atopic dermatitis and coronary artery disease. Additionally, we highlight current developments and discuss challenges to be faced in the future. Exemplary, we point to residual challenges in detecting disease-relevant variants resulting from difficulties in the interpretation of candidate variants and their potential interactions. While polygenic risk scores represent promising tools for the stratification of patient groups, currently, polygenic risk scores are not accurate enough for clinical setting. Precision medicine, incorporating additional data from genomics, transcriptomics and proteomics experiments, may enable the identification of distinct disease pathogeneses. In the future, data-intensive biomedical innovation will hopefully lead to improved patient stratification for personalized medicine.

Criteria for evaluating risk prediction of multiple outcomes

Risk prediction models have been developed in many contexts to classify individuals according to a single outcome, such as risk of a disease. Emerging “-omic” biomarkers provide panels of features that can simultaneously predict multiple outcomes from a single biological sample, creating issues of multiplicity reminiscent of exploratory hypothesis testing. Here I propose definitions of some basic criteria for evaluating prediction models of multiple outcomes. I define calibration in the multivariate setting and then distinguish between outcome-wise and individual-wise prediction, and within the latter between joint and panel-wise prediction. I give examples such as screening and early detection in which different senses of prediction may be more appropriate. In each case I propose definitions of sensitivity, specificity, concordance, positive and negative predictive value and relative utility. I link the definitions through a multivariate probit model, showing that the accuracy of a multivariate prediction model can be summarised by its covariance with a liability vector. I illustrate the concepts on a biomarker panel for early detection of eight cancers, and on polygenic risk scores for six common diseases.

Genetics and epigenetics of IBD

Inflammatory bowel diseases (IBD) are chronic intermittent inflammatory disorders of the gastrointestinal tract of unknown etiology but a clear genetic predisposition. Prompted by the first investigations on IBD families and twins, the genetic and epigenetic studies have produced an unprecedented amount of information in comparison with other immune-mediated or complex diseases. New inflammatory pathways and possible mechanisms of action have been disclosed, potentially leading to new-targeted therapy. However, the identification of genetic markers due to the great disease heterogeneity and the overwhelming contribution of environmental risk factors has not modified yet the disease management. The possibility for the future of a better prediction of disease course, response to therapy and therapy-related adverse events may allow a more efficient and personalized strategy. This review will focus on more recent discoveries that may potentially be of relevance in daily clinical practice.

Identifying Patients With Inflammatory Bowel Diseases at High vs Low Risk of Complications

People with Crohn's disease and ulcerative colitis have varying presentations and clinical consequences of their disease. Patients commonly ask about their prognosis, and what this diagnosis means for them. They are asking their clinicians to predict the future. The importance of predicting the course of any disease is to guide patient expectations and to guide treatment decisions. In the past decade the strategy of inflammatory bowel disease (IBD) treatment has shifted to treat patients earlier in the course of their disease, before irreversible damage occurs. Treatment approaches for disease categorized as mild, moderate or severe has most often been based on a current assessment of symptoms or disease activity without including a longitudinal assessment of a patient's disease course including past disease complications and surgeries. While a patient's current disease activity most typically drives these treatment decisions, optimally, treatment decisions would be made accounting for past disease activity and complications and the predicted future disease course. When developing a treatment plan for an individual patient, the immediate goal is to treat the current disease activity for relief of symptoms, and the long-term goal is to prevent progression of their disease due to complications. Since not all patients will progress to a complicated disease course, it is important to be able to select the right patients for the right therapy. Therefore, developing methods of stratifying patients into low-risk versus high-risk of complications will be an important aspect of treating IBD now and in the future.

Validity of polygenic risk scores: are we measuring what we think we are?

Polygenic risk scores (PRSs) have become the standard for quantifying genetic liability in the prediction of disease risks. PRSs are generally constructed as weighted sum scores of risk alleles using effect sizes from genome-wide association studies as their weights. The construction of PRSs is being improved with more appropriate selection of independent single-nucleotide polymorphisms (SNPs) and optimized estimation of their weights but is rarely reflected upon from a theoretical perspective, focusing on the validity of the risk score. Borrowing from psychometrics, this paper discusses the validity of PRSs and introduces the three main types of validity that are considered in the evaluation of tests and measurements: construct, content, and criterion validity. This introduction is followed by a discussion of three topics that challenge the validity of PRS, namely, their claimed independence of clinical risk factors, the consequences of relaxing SNP inclusion thresholds and the selection of SNP weights. This discussion of the validity of PRS reminds us that we need to keep questioning if weighted sums of risk alleles are measuring what we think they are in the various scenarios in which PRSs are used and that we need to keep exploring alternative modeling strategies that might better reflect the underlying biological pathways.

Comparative performances of machine learning methods for classifying Crohn Disease patients using genome-wide genotyping data

Crohn Disease (CD) is a complex genetic disorder for which more than 140 genes have been identified using genome wide association studies (GWAS). However, the genetic architecture of the trait remains largely unknown. The recent development of machine learning (ML) approaches incited us to apply them to classify healthy and diseased people according to their genomic information. The Immunochip dataset containing 18,227 CD patients and 34,050 healthy controls enrolled and genotyped by the international Inflammatory Bowel Disease genetic consortium (IIBDGC) has been re-analyzed using a set of ML methods: penalized logistic regression (LR), gradient boosted trees (GBT) and artificial neural networks (NN). The main score used to compare the methods was the Area Under the ROC Curve (AUC) statistics. The impact of quality control (QC), imputing and coding methods on LR results showed that QC methods and imputation of missing genotypes may artificially increase the scores. At the opposite, neither the patient/control ratio nor marker preselection or coding strategies significantly affected the results. LR methods, including Lasso, Ridge and ElasticNet provided similar results with a maximum AUC of 0.80. GBT methods like XGBoost, LightGBM and CatBoost, together with dense NN with one or more hidden layers, provided similar AUC values, suggesting limited epistatic effects in the genetic architecture of the trait. ML methods detected near all the genetic variants previously identified by GWAS among the best predictors plus additional predictors with lower effects. The robustness and complementarity of the different methods are also studied. Compared to LR, non-linear models such as GBT or NN may provide robust complementary approaches to identify and classify genetic markers.

The Role of Genetic Factors in Characterizing Extra-Intestinal Manifestations in Crohn's Disease Patients: Are Bayesian Machine Learning Methods Improving Outcome Predictions?

(1) Background: The high heterogeneity of inflammatory bowel disease (IBD) makes the study of this condition challenging. In subjects affected by Crohn’s disease (CD), extra-intestinal manifestations (EIMs) have a remarkable potential impact on health status. Increasing numbers of patient characteristics and the small size of analyzed samples make EIMs prediction very difficult. Under such constraints, Bayesian machine learning techniques (BMLTs) have been proposed as a robust alternative to classical models for outcome prediction. This study aims to determine whether BMLT could improve EIM prediction and statistical support for the decision-making process of clinicians. (2) Methods: Three of the most popular BMLTs were employed in this study: Naϊve Bayes (NB), Bayesian Network (BN) and Bayesian Additive Regression Trees (BART). They were applied to a retrospective observational Italian study of IBD genetics. (3) Results: The performance of the model is strongly affected by the features of the dataset, and BMLTs poorly classify EIM appearance. (4) Conclusions: This study shows that BMLTs perform worse than expected in classifying the presence of EIMs compared to classical statistical tools in a context where mixed genetic and clinical data are available but relevant data are also missing, as often occurs in clinical practice.

Molecular Profiling of Inflammatory Bowel Disease: Is It Ready for Use in Clinical Decision-Making?

Inflammatory bowel disease (IBD) is a heterogeneous disorder in terms of age at onset, clinical phenotypes, severity, disease course, and response to therapy. This underlines the need for predictive and precision medicine that can optimize diagnosis and disease management, provide more cost-effective strategies, and minimize the risk of adverse events. Ideally, we can leverage molecular profiling to predict the risk to develop IBD and disease progression. Despite substantial successes of genome-wide association studies in the identification of genetic variants affecting IBD susceptibility, molecular profiling of disease onset and progression as well as of treatment responses has lagged behind. Still, thanks to technological advances and good study designs, predicting phenotypes using genomics and transcriptomics in IBD has been rapidly evolving. In this review, we summarize the current status of prediction of disease risk, clinical course, and response to therapy based on clinical case presentations. We also discuss the potential and limitations of the currently used approaches.

A hybrid neural network model for predicting kidney disease in hypertension patients based on electronic health records

Background

Disease prediction based on Electronic Health Records (EHR) has become one hot research topic in biomedical community. Existing work mainly focuses on the prediction of one target disease, and little work is proposed for multiple associated diseases prediction. Meanwhile, a piece of EHR usually contains two main information: the textual description and physical indicators. However, existing work largely adopts statistical models with discrete features from numerical physical indicators in EHR, and fails to make full use of textual description information.

Methods

In this paper, we study the problem of kidney disease prediction in hypertension patients by using neural network model. Specifically, we first model the prediction problem as a binary classification task. Then we propose a hybrid neural network which incorporates Bidirectional Long Short-Term Memory (BiLSTM) and Autoencoder networks to fully capture the information in EHR.

Results

We construct a dataset based on a large number of raw EHR data. The dataset consists of totally 35,332 records from hypertension patients. Experimental results show that the proposed neural model achieves 89.7% accuracy for the task.

Conclusions

A hybrid neural network model was presented. Based on the constructed dataset, the comparison results of different models demonstrated the effectiveness of the proposed neural model. The proposed model outperformed traditional statistical models with discrete features and neural baseline systems.