AI-assisted discovery of an ethnicity-influenced driver of cell transformation in esophageal and gastroesophageal junction adenocarcinomas

Ghosh, Pradipta; Campos, Vinícius Jardim; Vo, Daniella T.; Guccione, Caitlin; Goheen-Holland, Vanae; Tindle, Courtney; Mazzini, Guilherme S.; He, Yudou; Alexandrov, Ludmil B.; Lippman, Scott M.; Gurski, Richard R; Das, Soumita; Yadlapati, Rena; Curtius, Kit; Sahoo, Debashis

doi:10.1172/jci.insight.161334

Cited by 7 publications

(6 citation statements)

References 39 publications

(72 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A method to build a network using transcriptomics data, in which gene clusters are connected by Boolean invariant relationships [43,44], has been used to track the progression of cellular states along any disease continuum. Previously, this methodology was used to identify translationally relevant cellular states with high degrees of accuracy in diverse samples and tissues [45][46][47][48][49][50][51][52][53][54][55][56][57][58]. Most recently, it was used to identify a therapeutic target to protect the gut barrier in inflammatory bowel disease [44].…”

Section: Boolean Implication Network Identifies Key Pathways In the P...mentioning

confidence: 99%

Molecular Signatures for Microbe-Associated Colorectal Cancers

Sayed,

Vo,

Alcantara

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

BackgroundGenetic factors and microbial imbalances play crucial roles in colorectal cancers (CRCs), yet the impact of infections on cancer initiation remains poorly understood. While bioinformatic approaches offer valuable insights, the rising incidence of CRCs creates a pressing need to precisely identify early CRC events. We constructed a network model to identify continuum states during CRC initiation spanning normal colonic tissue to pre-cancer lesions (adenomatous polyps) and examined the influence of microbes and host genetics.MethodsA Boolean network was built using a publicly available transcriptomic dataset from healthy and adenoma affected patients to identify an invariant Microbe-Associated Colorectal Cancer Signature (MACS). We focused onFusobacterium nucleatum(Fn), a CRC-associated microbe, as a model bacterium. MACS-associated genes and proteins were validated by RT-qPCR, RNA seq, ELISA, IF and IHCs in tissues and colon-derived organoids from genetically predisposed mice (CPC-APCMin+/-) and patients (FAP, Lynch Syndrome, PJS, and JPS).ResultsThe MACS that is upregulated in adenomas consists of four core genes/proteins: CLDN2/Claudin-2 (leakiness), LGR5/leucine-rich repeat-containing receptor (stemness), CEMIP/cell migration-inducing and hyaluronan-binding protein (epithelial-mesenchymal transition) and IL8/Interleukin-8 (inflammation). MACS was induced uponFninfection, but not in response to infection with other enteric bacteria or probiotics. MACS induction uponFninfection was higher inCPC-APCMin+/-organoids compared to WT controls. The degree of MACS expression in the patient-derived organoids (PDOs) generally corresponded with the known lifetime risk of CRCs.ConclusionsComputational prediction followed by validation in the organoid-based disease model identified the early events in CRC initiation. MACS reveals that the CRC-associated microbes induce a greater risk in the genetically predisposed hosts, suggesting its potential use for risk prediction and targeted cancer prevention.

show abstract

Section: Boolean Implication Network Identifies Key Pathways In the P...mentioning

confidence: 99%

Molecular Signatures for Microbe-Associated Colorectal Cancers

Sayed,

Vo,

Alcantara

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…BIRs represent all six types of possible gene associations-two symmetric and four asymmetric (17); the latter often go unrecognized by conventional methods. BIRs are presumed to reflect fundamental gene regulatory events in the continuum between health and disease (18)(19)(20)(21)(22). This approach has successfully been used to generate predictive models (a.k.a., disease maps) of sequential changes in gene expression patterns that are fulfilled by all samples (hence, invariant or universal) in the modeled disease (17,19,20,(22)(23)(24)(25)(26)(27).…”

Section: Study Design and Rationalementioning

confidence: 99%

“…BIRs are presumed to reflect fundamental gene regulatory events in the continuum between health and disease (18)(19)(20)(21)(22). This approach has successfully been used to generate predictive models (a.k.a., disease maps) of sequential changes in gene expression patterns that are fulfilled by all samples (hence, invariant or universal) in the modeled disease (17,19,20,(22)(23)(24)(25)(26)(27). These sequential changes, identified first in a training dataset and prioritized subsequently by machine learning, represent the myriads of continuum states between health and overt disease, manifesting as heterogeneous clinical presentations.…”

Section: Study Design and Rationalementioning

confidence: 99%

FORWARD: A Learning Framework for Logical Network Perturbations to Prioritize Targets for Drug Development

Sinha,

McLaren,

Mullick

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

Despite advances in artificial intelligence (AI), target-based drug development remains a costly, complex and imprecise process. We introduce F.O.R.W.A.R.D [Framework for Outcome-based Research and Drug Development], a network-based target prioritization approach and test its utility in the challenging therapeutic area of Inflammatory Bowel Diseases (IBD), which is a chronic condition of multifactorial origin. F.O.R.W.A.R.D leverages real-world outcomes, using a machine-learning classifier trained on transcriptomic data from seven prospective randomized clinical trials involving four drugs. It establishes a molecular signature of remission as the therapeutic goal and computes, by integrating principles of network connectivity, the likelihood that a drug’s action on its target(s) will induce the remission-associated genes. Benchmarking F.O.R.W.A.R.D against 210 completed clinical trials on 52 targets showed a perfect predictive accuracy of 100%. The success of F.O.R.W.A.R.D was achieved despite differences in targets, mechanisms, and trial designs. F.O.R.W.A.R.D-driven in-silico phase ’0’ trials revealed its potential to inform trial design, justify re-trialing failed drugs, and guide early terminations. With its extendable applications to other therapeutic areas and its iterative refinement with emerging trials, F.O.R.W.A.R.D holds the promise to transform drug discovery by generating foresight from hindsight and impacting research and development as well as human-in-the-loop clinical decision-making.

show abstract

“…Another development which may significantly impact cancer risk prediction is the inclusion of genomic factors as part of risk assessment. At present, the risk prediction algorithms which have been developed for use in primary care do not include genetic or genomic markers associated with oesophageal cancer because these are not yet routinely tested for or recorded in primary care [60][61][62]. However, understanding the evolution of oesophageal tumours over time, and being able to integrate successive genomic, histological and clinical records may drive the development of risk prediction algorithms for oesophageal cancer that incorporate genetic and genomic risk factors [12].…”

Section: Refining Predictive Risk Algorithms To Include Genomic Risk ...mentioning

confidence: 99%

Ethical and legal implications of implementing risk algorithms for early detection and screening for oesophageal cancer, now and in the future

Brigden,

Mitchell,

Redrup Hill

et al. 2023

PLoS ONE

View full text Add to dashboard Cite

Background Oesophageal cancer has significant morbidity and mortality but late diagnosis is common since early signs of disease are frequently misinterpreted. Project DELTA aims to enable earlier detection and treatment through targeted screening using a novel risk prediction algorithm for oesophageal cancer (incorporating risk factors of Barrett’s oesophagus including prescriptions for acid-reducing medications (CanPredict)), together with a non-invasive, low-cost sampling device (CytospongeTM). However, there are many barriers to implementation, and this paper identifies key ethical and legal challenges to implementing these personalised prevention strategies for Barrett’s oesophagus/oesophageal cancer. Methods To identify ethical and legal issues relevant to the deployment of a risk prediction tool for oesophageal cancer into primary care, we adopted an interdisciplinary approach, incorporating targeted informal literature reviews, interviews with expert collaborators, a multidisciplinary workshop and ethical and legal analysis. Results Successful implementation raises many issues including ensuring transparency and effective risk communication; addressing bias and inequity; managing resources appropriately and avoiding exceptionalism. Clinicians will need support and training to use cancer risk prediction algorithms, ensuring that they understand how risk algorithms supplement rather than replace medical decision-making. Workshop participants had concerns about liability for harms arising from risk algorithms, including from potential bias and inequitable implementation. Determining strategies for risk communication enabling transparency but avoiding exceptionalist approaches are a significant challenge. Future challenges include using artificial intelligence to bolster risk assessment, incorporating genomics into risk tools, and deployment by non-health professional users. However, these strategies could improve detection and outcomes. Conclusions Novel pathways incorporating risk prediction algorithms hold considerable promise, especially when combined with low-cost sampling. However immediate priorities should be to develop risk communication strategies that take account of using validated risk algorithms, and to ensure equitable implementation. Resolving questions about liability for harms arising should be a longer-term objective.

show abstract

AI-assisted discovery of an ethnicity-influenced driver of cell transformation in esophageal and gastroesophageal junction adenocarcinomas

Cited by 7 publications

References 39 publications

Molecular Signatures for Microbe-Associated Colorectal Cancers

Molecular Signatures for Microbe-Associated Colorectal Cancers

FORWARD: A Learning Framework for Logical Network Perturbations to Prioritize Targets for Drug Development

Ethical and legal implications of implementing risk algorithms for early detection and screening for oesophageal cancer, now and in the future

Contact Info

Product

Resources

About