Combinatorial prediction of therapeutic perturbations using causally-inspired neural networks.

As an alternative to target-driven drug discovery, phenotype-driven approaches identify compounds that counteract the overall disease effects by analyzing phenotypic signatures. Our study introduces a novel approach to this field, aiming to expand the search space for new therapeutic agents. We introduce PDGrapher, a causally-inspired graph neural network model designed to predict arbitrary perturbagens - sets of therapeutic targets - capable of reversing disease effects. Unlike existing methods that learn responses to perturbations, PDGrapher solves the inverse problem, which is to infer the perturbagens necessary to achieve a specific response - i.e., directly predicting perturbagens by learning which perturbations elicit a desired response. Experiments across eight datasets of genetic and chemical perturbations show that PDGrapher successfully predicted effective perturbagens in up to 9% additional test samples and ranked therapeutic targets up to 35% higher than competing methods. A key innovation of PDGrapher is its direct prediction capability, which contrasts with the indirect, computationally intensive models traditionally used in phenotypedriven drug discovery that only predict changes in phenotypes due to perturbations. The direct approach enables PDGrapher to train up to 30 times faster, representing a significant leap in efficiency. Our results suggest that PDGrapher can advance phenotype-driven drug discovery, offering a fast and comprehensive approach to identifying therapeutically useful perturbations.

Crosstalk with lung fibroblasts shapes the growth and therapeutic response of mesothelioma cells.

Mesothelioma is an aggressive cancer of the mesothelial layer associated with an extensive fibrotic response. The latter is in large part mediated by cancer-associated fibroblasts which mediate tumour progression and poor prognosis. However, understanding of the crosstalk between cancer cells and fibroblasts in this disease is mostly lacking. Here, using co-cultures of patient-derived mesothelioma cell lines and lung fibroblasts, we demonstrate that fibroblast activation is a self-propagated process producing a fibrotic extracellular matrix (ECM) and triggering drug resistance in mesothelioma cells. Following characterisation of mesothelioma cells/fibroblasts signalling crosstalk, we identify several FDA-approved targeted therapies as far more potent than standard-of-care Cisplatin/Pemetrexed in ECM-embedded co-culture spheroid models. In particular, the SRC family kinase inhibitor, Saracatinib, extends overall survival well beyond standard-of-care in a mesothelioma genetically-engineered mouse model. In short, we lay the foundation for the rational design of novel therapeutic strategies targeting mesothelioma/fibroblast communication for the treatment of mesothelioma patients.

Genomic-driven nutritional interventions for radiotherapy-resistant rectal cancer patient.

Radiotherapy response of rectal cancer patients is dependent on a myriad of molecular mechanisms including response to stress, cell death, and cell metabolism. Modulation of lipid metabolism emerges as a unique strategy to improve radiotherapy outcomes due to its accessibility by bioactive molecules within foods. Even though a few radioresponse modulators have been identified using experimental techniques, trying to experimentally identify all potential modulators is intractable. Here we introduce a machine learning (ML) approach to interrogate the space of bioactive molecules within food for potential modulators of radiotherapy response and provide phytochemically-enriched recipes that encapsulate the benefits of discovered radiotherapy modulators. Potential radioresponse modulators were identified using a genomic-driven network ML approach, metric learning and domain knowledge. Then, recipes from the Recipe1M database were optimized to provide ingredient substitutions maximizing the number of predicted modulators whilst preserving the recipe's culinary attributes. This work provides a pipeline for the design of genomic-driven nutritional interventions to improve outcomes of rectal cancer patients undergoing radiotherapy.

An AI-powered patient triage platform for future viral outbreaks using COVID-19 as a disease model.

Over the last century, outbreaks and pandemics have occurred with disturbing regularity, necessitating advance preparation and large-scale, coordinated response. Here, we developed a machine learning predictive model of disease severity and length of hospitalization for COVID-19, which can be utilized as a platform for future unknown viral outbreaks. We combined untargeted metabolomics on plasma data obtained from COVID-19 patients (n = 111) during hospitalization and healthy controls (n = 342), clinical and comorbidity data (n = 508) to build this patient triage platform, which consists of three parts: (i) the clinical decision tree, which amongst other biomarkers showed that patients with increased eosinophils have worse disease prognosis and can serve as a new potential biomarker with high accuracy (AUC = 0.974), (ii) the estimation of patient hospitalization length with ± 5 days error (R2 = 0.9765) and (iii) the prediction of the disease severity and the need of patient transfer to the intensive care unit. We report a significant decrease in serotonin levels in patients who needed positive airway pressure oxygen and/or were intubated. Furthermore, 5-hydroxy tryptophan, allantoin, and glucuronic acid metabolites were increased in COVID-19 patients and collectively they can serve as biomarkers to predict disease progression. The ability to quickly identify which patients will develop life-threatening illness would allow the efficient allocation of medical resources and implementation of the most effective medical interventions. We would advocate that the same approach could be utilized in future viral outbreaks to help hospitals triage patients more effectively and improve patient outcomes while optimizing healthcare resources.

Alzheimer's disease: using gene/protein network machine learning for molecule discovery in olive oil.

Alzheimer's disease (AD) poses a profound human, social, and economic burden. Previous studies suggest that extra virgin olive oil (EVOO) may be helpful in preventing cognitive decline. Here, we present a network machine learning method for identifying bioactive phytochemicals in EVOO with the highest potential to impact the protein network linked to the development and progression of the AD. A balanced classification accuracy of 70.3 ± 2.6% was achieved in fivefold cross-validation settings for predicting late-stage experimental drugs targeting AD from other clinically approved drugs. The calibrated machine learning algorithm was then used to predict the likelihood of existing drugs and known EVOO phytochemicals to be similar in action to the drugs impacting AD protein networks. These analyses identified the following ten EVOO phytochemicals with the highest likelihood of being active against AD: quercetin, genistein, luteolin, palmitoleate, stearic acid, apigenin, epicatechin, kaempferol, squalene, and daidzein (in the order from the highest to the lowest likelihood). This in silico study presents a framework that brings together artificial intelligence, analytical chemistry, and omics studies to identify unique therapeutic agents. It provides new insights into how EVOO constituents may help treat or prevent AD and potentially provide a basis for consideration in future clinical studies.

Deep learning to detect macular atrophy in wet age-related macular degeneration using optical coherence tomography.

Here, we have developed a deep learning method to fully automatically detect and quantify six main clinically relevant atrophic features associated with macular atrophy (MA) using optical coherence tomography (OCT) analysis of patients with wet age-related macular degeneration (AMD). The development of MA in patients with AMD results in irreversible blindness, and there is currently no effective method of early diagnosis of this condition, despite the recent development of unique treatments. Using OCT dataset of a total of 2211 B-scans from 45 volumetric scans of 8 patients, a convolutional neural network using one-against-all strategy was trained to present all six atrophic features followed by a validation to evaluate the performance of the models. The model predictive performance has achieved a mean dice similarity coefficient score of 0.706 ± 0.039, a mean Precision score of 0.834 ± 0.048, and a mean Sensitivity score of 0.615 ± 0.051. These results show the unique potential of using artificially intelligence-aided methods for early detection and identification of the progression of MA in wet AMD, which can further support and assist clinical decisions.

Treatment of wild-type mice with 2,3-butanediol, a urinary biomarker of Fmo5 -/- mice, decreases plasma cholesterol and epididymal fat deposition.

We previously showed that Fmo5 -/- mice exhibit a lean phenotype and slower metabolic ageing. Their characteristics include lower plasma glucose and cholesterol, greater glucose tolerance and insulin sensitivity, and a reduction in age-related weight gain and whole-body fat deposition. In this paper, nuclear magnetic resonance (NMR) spectroscopy-based metabolite analyses of the urine of Fmo5 -/- and wild-type mice identified two isomers of 2,3-butanediol as discriminating urinary biomarkers of Fmo5 -/- mice. Antibiotic-treatment of Fmo5 -/- mice increased plasma cholesterol concentration and substantially reduced urinary excretion of 2,3-butanediol isomers, indicating that the gut microbiome contributed to the lower plasma cholesterol of Fmo5 -/- mice, and that 2,3-butanediol is microbially derived. Short- and long-term treatment of wild-type mice with a 2,3-butanediol isomer mix decreased plasma cholesterol and epididymal fat deposition but had no effect on plasma concentrations of glucose or insulin, or on body weight. In the case of long-term treatment, the effects were maintained after withdrawal of 2,3-butanediol. Short-, but not long-term treatment, also decreased plasma concentrations of triglycerides and non-esterified fatty acids. Fecal transplant from Fmo5 -/- to wild-type mice had no effect on plasma cholesterol, and 2,3-butanediol was not detected in the urine of recipient mice, suggesting that the microbiota of the large intestine was not the source of 2,3-butanediol. However, 2,3-butanediol was detected in the stomach of Fmo5 -/- mice, which was enriched for Lactobacillus genera, known to produce 2,3-butanediol. Our results indicate a microbial contribution to the phenotypic characteristic of Fmo5 -/- mice of decreased plasma cholesterol and identify 2,3-butanediol as a potential agent for lowering plasma cholesterol.

Lipidomics and Redox Lipidomics Indicate Early Stage Alcohol-Induced Liver Damage.

Alcoholic fatty liver disease (AFLD) is characterized by lipid accumulation and inflammation and can progress to cirrhosis and cancer in the liver. AFLD diagnosis currently relies on histological analysis of liver biopsies. Early detection permits interventions that would prevent progression to cirrhosis or later stages of the disease. Herein, we have conducted the first comprehensive time-course study of lipids using novel state-of-the art lipidomics methods in plasma and liver in the early stages of a mouse model of AFLD, i.e., Lieber-DeCarli diet model. In ethanol-treated mice, changes in liver tissue included up-regulation of triglycerides (TGs) and oxidized TGs and down-regulation of phosphatidylcholine, lysophosphatidylcholine, and 20-22-carbon-containing lipid-mediator precursors. An increase in oxidized TGs preceded histological signs of early AFLD, i.e., steatosis, with these changes observed in both the liver and plasma. The major lipid classes dysregulated by ethanol play important roles in hepatic inflammation, steatosis, and oxidative damage. Conclusion: Alcohol consumption alters the liver lipidome before overt histological markers of early AFLD. This introduces the exciting possibility that specific lipids may serve as earlier biomarkers of AFLD than those currently being used.

Metabonomics study of the effects of single copy mutant KRAS in the presence or absence of WT allele using human HCT116 isogenic cell lines.

INTRODUCTION: KRAS was one of the earliest human oncogenes to be described and is one of the most commonly mutated genes in different human cancers, including colorectal cancer. Despite KRAS mutants being known driver mutations, KRAS has proved difficult to target therapeutically, necessitating a comprehensive understanding of the molecular mechanisms underlying KRAS-driven cellular transformation. OBJECTIVES: To investigate the metabolic signatures associated with single copy mutant KRAS in isogenic human colorectal cancer cells and to determine what metabolic pathways are affected. METHODS: Using NMR-based metabonomics, we compared wildtype (WT)-KRAS and mutant KRAS effects on cancer cell metabolism using metabolic profiling of the parental KRAS G13D/+ HCT116 cell line and its isogenic, derivative cell lines KRAS +/- and KRAS G13D/-. RESULTS: Mutation in the KRAS oncogene leads to a general metabolic remodelling to sustain growth and counter stress, including alterations in the metabolism of amino acids and enhanced glutathione biosynthesis. Additionally, we show that KRASG13D/+ and KRASG13D/- cells have a distinct metabolic profile characterized by dysregulation of TCA cycle, up-regulation of glycolysis and glutathione metabolism pathway as well as increased glutamine uptake and acetate utilization. CONCLUSIONS: Our study showed the effect of a single point mutation in one KRAS allele and KRAS allele loss in an isogenic genetic background, hence avoiding confounding genetic factors. Metabolic differences among different KRAS mutations might play a role in their different responses to anticancer treatments and hence could be exploited as novel metabolic vulnerabilities to develop more effective therapies against oncogenic KRAS.

Virus-induced Volatile Organic Compounds Are Detectable in Exhaled Breath during Pulmonary Infection.

Rationale: Chronic obstructive pulmonary disease (COPD) is a condition punctuated by acute exacerbations commonly triggered by viral and/or bacterial infection. Early identification of exacerbation triggers is important to guide appropriate therapy, but currently available tests are slow and imprecise. Volatile organic compounds (VOCs) can be detected in exhaled breath and have the potential to be rapid tissue-specific biomarkers of infection etiology. Objectives: To determine whether volatile organic compound measurement could distinguish viral from bacterial infection in COPD. Methods: We used serial sampling within in vitro and in vivo studies to elucidate the dynamic changes that occur in VOC production during acute respiratory viral infection. Highly sensitive gas chromatography-mass spectrometry techniques were used to measure VOC production from infected airway epithelial-cell cultures and in exhaled breath samples from healthy subjects experimentally challenged with rhinovirus (RV)-A16 and from subjects with COPD with naturally occurring exacerbations. Measurements and Main Results: We identified a novel VOC signature comprising decane and other long-chain alkane compounds that is induced during RV infection of cultured airway epithelial cells and is also increased in the exhaled breath from healthy subjects experimentally challenged with RV and from patients with COPD during naturally occurring viral exacerbations. These compounds correlated with the magnitude of antiviral immune responses, viral burden, and exacerbation severity but were not induced by bacterial infection, suggesting that they represent a specific virus-inducible signature. Conclusions: Our study highlights the potential for measurement of exhaled breath VOCs as rapid, noninvasive biomarkers of viral infection. Further studies are needed to determine whether measurement of these signatures could be used to guide more targeted therapy with antibiotic/antiviral agents for COPD exacerbations.

Phytochemically rich dietary components and the risk of colorectal cancer: A systematic review and meta-analysis of observational studies.

BACKGROUND: Personalized nutrition and protective diets and lifestyles represent a key cancer research priority. The association between consumption of specific dietary components and colorectal cancer (CRC) incidence has been evaluated by a number of population-based studies, which have identified certain food items as having protective potential, though the findings have been inconsistent. Herein we present a systematic review and meta-analysis on the potential protective role of five common phytochemically rich dietary components (nuts, cruciferous vegetables, citrus fruits, garlic and tomatoes) in reducing CRC risk. AIM: To investigate the independent impact of increased intake of specific dietary constituents on CRC risk in the general population. METHODS: Medline and Embase were systematically searched, from time of database inception to January 31, 2020, for observational studies reporting CRC incidence relative to intake of one or more of nuts, cruciferous vegetables, citrus fruits, garlic and/or tomatoes in the general population. Data were extracted by two independent reviewers and analyzed in accordance with the Meta-analysis of Observational Studies in Epidemiology (MOOSE) and Preferred Reporting Items for Systematic Review and Meta-analysis (PRISMA) reporting guidelines and according to predefined inclusion/exclusion criteria. Effect sizes of studies were pooled using a random-effects model. RESULTS: Forty-six studies were identified. CRC risk was significantly reduced in patients with higher vs lower consumption of cruciferous vegetables [odds ratio (OR) = 0.90; 95% confidence interval (CI): 0.85-0.95; P < 0.005], citrus fruits (OR = 0.90; 95%CI: 0.84-0.96; P < 0.005), garlic (OR = 0.83; 95%CI: 0.76-0.91; P < 0.005) and tomatoes (OR = 0.89; 95%CI: 0.84-0.95; P < 0.005). Subgroup analysis showed that this association sustained when looking at case-control studies alone, for all of these four food items, but no significant difference was found in analysis of cohort studies alone. Nut consumption exhibited a similar trend, but overall results were not significant (OR = 0.72; 95%CI: 0.50-1.03; P < 0.07; I 2 = 90.70%). Putative anticarcinogenic mechanisms are proposed using gene-set enrichment analysis of gene/protein perturbations caused by active compounds within each food item. CONCLUSION: Increased cruciferous vegetable, garlic, citrus fruit and tomato consumption are all inversely associated with CRC risk. These findings highlight the potential for developing precision nutrition strategies for CRC prevention.

Predicting anticancer hyperfoods with graph convolutional networks.

BACKGROUND: Recent efforts in the field of nutritional science have allowed the discovery of disease-beating molecules within foods based on the commonality of bioactive food molecules to FDA-approved drugs. The pioneering work in this field used an unsupervised network propagation algorithm to learn the systemic-wide effect on the human interactome of 1962 FDA-approved drugs and a supervised algorithm to predict anticancer therapeutics using the learned representations. Then, a set of bioactive molecules within foods was fed into the model, which predicted molecules with cancer-beating potential.The employed methodology consisted of disjoint unsupervised feature generation and classification tasks, which can result in sub-optimal learned drug representations with respect to the classification task. Additionally, due to the disjoint nature of the tasks, the employed approach proved cumbersome to optimize, requiring testing of thousands of hyperparameter combinations and significant computational resources.To overcome the technical limitations highlighted above, we represent each drug as a graph (human interactome) with its targets as binary node features on the graph and formulate the problem as a graph classification task. To solve this task, inspired by the success of graph neural networks in graph classification problems, we use an end-to-end graph neural network model operating directly on the graphs, which learns drug representations to optimize model performance in the prediction of anticancer therapeutics. RESULTS: The proposed model outperforms the baseline approach in the anticancer therapeutic prediction task, achieving an F1 score of 67.99%±2.52% and an AUPR of 73.91%±3.49%. It is also shown that the model is able to capture knowledge of biological pathways to predict anticancer molecules based on the molecules' effects on cancer-related pathways. CONCLUSIONS: We introduce an end-to-end graph convolutional model to predict cancer-beating molecules within food. The introduced model outperforms the existing baseline approach, and shows interpretability, paving the way to the future of a personalized nutritional science approach allowing the development of nutrition strategies for cancer prevention and/or therapeutics.

Network machine learning maps phytochemically rich "Hyperfoods" to fight COVID-19.

In this paper, we introduce a network machine learning method to identify potential bioactive anti-COVID-19 molecules in foods based on their capacity to target the SARS-CoV-2-host gene-gene (protein-protein) interactome. Our analyses were performed using a supercomputing DreamLab App platform, harnessing the idle computational power of thousands of smartphones. Machine learning models were initially calibrated by demonstrating that the proposed method can predict anti-COVID-19 candidates among experimental and clinically approved drugs (5658 in total) targeting COVID-19 interactomics with the balanced classification accuracy of 80-85% in 5-fold cross-validated settings. This identified the most promising drug candidates that can be potentially "repurposed" against COVID-19 including common drugs used to combat cardiovascular and metabolic disorders, such as simvastatin, atorvastatin and metformin. A database of 7694 bioactive food-based molecules was run through the calibrated machine learning algorithm, which identified 52 biologically active molecules, from varied chemical classes, including flavonoids, terpenoids, coumarins and indoles predicted to target SARS-CoV-2-host interactome networks. This in turn was used to construct a "food map" with the theoretical anti-COVID-19 potential of each ingredient estimated based on the diversity and relative levels of candidate compounds with antiviral properties. We expect this in silico predicted food map to play an important role in future clinical studies of precision nutrition interventions against COVID-19 and other viral diseases.

De Novo Lipogenesis Alters the Phospholipidome of Esophageal Adenocarcinoma.

The incidence of esophageal adenocarcinoma is rising, survival remains poor, and new tools to improve early diagnosis and precise treatment are needed. Cancer phospholipidomes quantified with mass spectrometry imaging (MSI) can support objective diagnosis in minutes using a routine frozen tissue section. However, whether MSI can objectively identify primary esophageal adenocarcinoma is currently unknown and represents a significant challenge, as this microenvironment is complex with phenotypically similar tissue-types. Here, we used desorption electrospray ionization-MSI (DESI-MSI) and bespoke chemometrics to assess the phospholipidomes of esophageal adenocarcinoma and relevant control tissues. Multivariate models derived from phospholipid profiles of 117 patients were highly discriminant for esophageal adenocarcinoma both in discovery (AUC = 0.97) and validation cohorts (AUC = 1). Among many other changes, esophageal adenocarcinoma samples were markedly enriched for polyunsaturated phosphatidylglycerols with longer acyl chains, with stepwise enrichment in premalignant tissues. Expression of fatty acid and glycerophospholipid synthesis genes was significantly upregulated, and characteristics of fatty acid acyls matched glycerophospholipid acyls. Mechanistically, silencing the carbon switch ACLY in esophageal adenocarcinoma cells shortened glycerophospholipid chains, linking de novo lipogenesis to the phospholipidome. Thus, DESI-MSI can objectively identify invasive esophageal adenocarcinoma from a number of premalignant tissues and unveils mechanisms of phospholipidomic reprogramming. SIGNIFICANCE: These results call for accelerated diagnosis studies using DESI-MSI in the upper gastrointestinal endoscopy suite, as well as functional studies to determine how polyunsaturated phosphatidylglycerols contribute to esophageal carcinogenesis.

Metabolic characterization of colorectal cancer cells harbouring different KRAS mutations in codon 12, 13, 61 and 146 using human SW48 isogenic cell lines.

INTRODUCTION: Kirsten Rat Sarcoma Viral Oncogene Homolog (KRAS) mutations occur in approximately one-third of colorectal (CRC) tumours and have been associated with poor prognosis and resistance to some therapeutics. In addition to the well-documented pro-tumorigenic role of mutant Ras alleles, there is some evidence suggesting that not all KRAS mutations are equal and the position and type of amino acid substitutions regulate biochemical activity and transforming capacity of KRAS mutations. OBJECTIVES: To investigate the metabolic signatures associated with different KRAS mutations in codons 12, 13, 61 and 146 and to determine what metabolic pathways are affected by different KRAS mutations. METHODS: We applied an NMR-based metabonomics approach to compare the metabolic profiles of the intracellular extracts and the extracellular media from isogenic human SW48 CRC cell lines with different KRAS mutations in codons 12 (G12D, G12A, G12C, G12S, G12R, G12V), 13 (G13D), 61 (Q61H) and 146 (A146T) with their wild-type counterpart. We used false discovery rate (FDR)-corrected analysis of variance (ANOVA) to determine metabolites that were statistically significantly different in concentration between the different mutants. RESULTS: CRC cells carrying distinct KRAS mutations exhibited differential metabolic remodelling, including differences in glycolysis, glutamine utilization and in amino acid, nucleotide and hexosamine metabolism. CONCLUSIONS: Metabolic differences among different KRAS mutations might play a role in their different responses to anticancer treatments and hence could be exploited as novel metabolic vulnerabilities to develop more effective therapies against oncogenic KRAS.

Precision Medicine in Pancreatic Disease-Knowledge Gaps and Research Opportunities: Summary of a National Institute of Diabetes and Digestive and Kidney Diseases Workshop.

A workshop on research gaps and opportunities for Precision Medicine in Pancreatic Disease was sponsored by the National Institute of Diabetes and Digestive Kidney Diseases on July 24, 2019, in Pittsburgh. The workshop included an overview lecture on precision medicine in cancer and 4 sessions: (1) general considerations for the application of bioinformatics and artificial intelligence; (2) omics, the combination of risk factors and biomarkers; (3) precision imaging; and (4) gaps, barriers, and needs to move from precision to personalized medicine for pancreatic disease. Current precision medicine approaches and tools were reviewed, and participants identified knowledge gaps and research needs that hinder bringing precision medicine to pancreatic diseases. Most critical were (a) multicenter efforts to collect large-scale patient data sets from multiple data streams in the context of environmental and social factors; (b) new information systems that can collect, annotate, and quantify data to inform disease mechanisms; (c) novel prospective clinical trial designs to test and improve therapies; and (d) a framework for measuring and assessing the value of proposed approaches to the health care system. With these advances, precision medicine can identify patients early in the course of their pancreatic disease and prevent progression to chronic or fatal illness.

A Unified Conceptual Framework for Metabolic Phenotyping in Diagnosis and Prognosis.

Understanding metabotype (multicomponent metabolic characteristics) variation can help to generate new diagnostic and prognostic biomarkers, as well as models, with potential to impact on patient management. We present a suite of conceptual approaches for the generation, analysis, and understanding of metabotypes from body fluids and tissues. We describe and exemplify four fundamental approaches to the generation and utilization of metabotype data via multiparametric measurement of (i) metabolite levels, (ii) metabolic trajectories, (iii) metabolic entropies, and (iv) metabolic networks and correlations in space and time. This conceptual framework can underpin metabotyping in the scenario of personalized medicine, with the aim of improving clinical outcomes for patients, but the framework will have value and utility in areas of metabolic profiling well beyond this exemplar.

HyperFoods: Machine intelligent mapping of cancer-beating molecules in foods.

Recent data indicate that up-to 30-40% of cancers can be prevented by dietary and lifestyle measures alone. Herein, we introduce a unique network-based machine learning platform to identify putative food-based cancer-beating molecules. These have been identified through their molecular biological network commonality with clinically approved anti-cancer therapies. A machine-learning algorithm of random walks on graphs (operating within the supercomputing DreamLab platform) was used to simulate drug actions on human interactome networks to obtain genome-wide activity profiles of 1962 approved drugs (199 of which were classified as "anti-cancer" with their primary indications). A supervised approach was employed to predict cancer-beating molecules using these 'learned' interactome activity profiles. The validated model performance predicted anti-cancer therapeutics with classification accuracy of 84-90%. A comprehensive database of 7962 bioactive molecules within foods was fed into the model, which predicted 110 cancer-beating molecules (defined by anti-cancer drug likeness threshold of >70%) with expected capacity comparable to clinically approved anti-cancer drugs from a variety of chemical classes including flavonoids, terpenoids, and polyphenols. This in turn was used to construct a 'food map' with anti-cancer potential of each ingredient defined by the number of cancer-beating molecules found therein. Our analysis underpins the design of next-generation cancer preventative and therapeutic nutrition strategies.