Integrative analysis of multimodal patient data identifies personalized predictors of tuberculosis treatment prognosis.

Tuberculosis (TB) afflicted 10.6 million people in 2021, and its global burden is increasing due to multidrug-resistant TB (MDR-TB) and extensively resistant TB (XDR-TB). Here, we analyze multi-domain information from 5,060 TB patients spanning 10 countries with high burden of MDR-TB from the NIAID TB Portals database to determine predictors of TB treatment outcome. Our analysis revealed significant associations between radiological, microbiological, therapeutic, and demographic data modalities. Our machine learning model, built with 203 features across modalities outperforms models built using each modality alone in predicting treatment outcomes, with an accuracy of 83% and area under the curve of 0.84. Notably, our analysis revealed that the drug regimens Bedaquiline-Clofazimine-Cycloserine-Levofloxacin-Linezolid and Bedaquiline-Clofazimine-Linezolid-Moxifloxacin were associated with treatment success and failure, respectively, for MDR non-XDR-TB. Drug combinations predicted to be synergistic by the INDIGO algorithm performed better than antagonistic combinations. Our prioritized set of features predictive of treatment outcomes can ultimately guide the personalized clinical management of TB.

Leveraging metabolic modeling and machine learning to uncover modulators of quiescence depth.

Quiescence, a temporary withdrawal from the cell cycle, plays a key role in tissue homeostasis and regeneration. Quiescence is increasingly viewed as a continuum between shallow and deep quiescence, reflecting different potentials to proliferate. The depth of quiescence is altered in a range of diseases and during aging. Here, we leveraged genome-scale metabolic modeling (GEM) to define the metabolic and epigenetic changes that take place with quiescence deepening. We discovered contrasting changes in lipid catabolism and anabolism and diverging trends in histone methylation and acetylation. We then built a multi-cell type machine learning model that accurately predicts quiescence depth in diverse biological contexts. Using both machine learning and genome-scale flux simulations, we performed high-throughput screening of chemical and genetic modulators of quiescence and identified novel small molecule and genetic modulators with relevance to cancer and aging.

Metabolic stratification of human breast tumors reveal subtypes of clinical and therapeutic relevance.

Extensive metabolic heterogeneity in breast cancers has limited the deployment of metabolic therapies. To enable patient stratification, we studied the metabolic landscape in breast cancers (∼3000 patients combined) and identified three subtypes with increasing degrees of metabolic deregulation. Subtype M1 was found to be dependent on bile-acid biosynthesis, whereas M2 showed reliance on methionine pathway, and M3 engaged fatty-acid, nucleotide, and glucose metabolism. The extent of metabolic alterations correlated strongly with tumor aggressiveness and patient outcome. This pattern was reproducible in independent datasets and using in vivo tumor metabolite data. Using machine-learning, we identified robust and generalizable signatures of metabolic subtypes in tumors and cell lines. Experimental inhibition of metabolic pathways in cell lines representing metabolic subtypes revealed subtype-specific sensitivity, therapeutically relevant drugs, and promising combination therapies. Taken together, metabolic stratification of breast cancers can thus aid in predicting patient outcome and designing precision therapies.

Anthocyanin: A Natural Dye Extracted from Hibiscus Sabdariffa (L.) for Textile and Dye Industries.

Environmental pollution is one of the major issues facing all countries throughout the world. Environmental degradation is occurring and creating crises in day-to-day life due to the increasing amount of chemicals used in industries, where even the effluents processed out after treatment also contain some trace elements. Hence the extraction of enzymes using natural methods is an alternative for the production of dye in order to reduce pollution, which in turn helps to nourish and protect the environment for future generations. Hibiscus sabdariffa (L.) is a rich source of anthocyanins that is further enhanced by callus formation and accumulated by increasing the sucrose concentration. Anthocyanin pigments were extracted using acidified ethanol. The dye obtained was screened by GC-MS analysis and its dyeing process used in the textile industry. The study showed certain properties affected the coloring nature depending on the cloth used. The color of anthocyanin pigment depends on the pH maintained and also shows adaptability to varied environmental conditions.

TaxiBGC: a Taxonomy-Guided Approach for Profiling Experimentally Characterized Microbial Biosynthetic Gene Clusters and Secondary Metabolite Production Potential in Metagenomes.

Biosynthetic gene clusters (BGCs) in microbial genomes encode bioactive secondary metabolites (SMs), which can play important roles in microbe-microbe and host-microbe interactions. Given the biological significance of SMs and the current profound interest in the metabolic functions of microbiomes, the unbiased identification of BGCs from high-throughput metagenomic data could offer novel insights into the complex chemical ecology of microbial communities. Currently available tools for predicting BGCs from shotgun metagenomes have several limitations, including the need for computationally demanding read assembly, predicting a narrow breadth of BGC classes, and not providing the SM product. To overcome these limitations, we developed taxonomy-guided identification of biosynthetic gene clusters (TaxiBGC), a command-line tool for predicting experimentally characterized BGCs (and inferring their known SMs) in metagenomes by first pinpointing the microbial species likely to harbor them. We benchmarked TaxiBGC on various simulated metagenomes, showing that our taxonomy-guided approach could predict BGCs with much-improved performance (mean F1 score, 0.56; mean PPV score, 0.80) compared with directly identifying BGCs by mapping sequencing reads onto the BGC genes (mean F1 score, 0.49; mean PPV score, 0.41). Next, by applying TaxiBGC on 2,650 metagenomes from the Human Microbiome Project and various case-control gut microbiome studies, we were able to associate BGCs (and their SMs) with different human body sites and with multiple diseases, including Crohn's disease and liver cirrhosis. In all, TaxiBGC provides an in silico platform to predict experimentally characterized BGCs and their SM production potential in metagenomic data while demonstrating important advantages over existing techniques. IMPORTANCE Currently available bioinformatics tools to identify BGCs from metagenomic sequencing data are limited in their predictive capability or ease of use to even computationally oriented researchers. We present an automated computational pipeline called TaxiBGC, which predicts experimentally characterized BGCs (and infers their known SMs) in shotgun metagenomes by first considering the microbial species source. Through rigorous benchmarking techniques on simulated metagenomes, we show that TaxiBGC provides a significant advantage over existing methods. When demonstrating TaxiBGC on thousands of human microbiome samples, we associate BGCs encoding bacteriocins with different human body sites and diseases, thereby elucidating a possible novel role of this antibiotic class in maintaining the stability of microbial ecosystems throughout the human body. Furthermore, we report for the first time gut microbial BGC associations shared among multiple pathologies. Ultimately, we expect our tool to facilitate future investigations into the chemical ecology of microbial communities across diverse niches and pathologies.

Protocol for CAROM: A machine learning tool to predict post-translational regulation from metabolic signatures.

This protocol describes CAROM, a computational tool that combines genome-scale metabolic networks (GEMs) and machine learning to identify enzyme targets of post-translational modifications (PTMs). Condition-specific enzyme and reaction properties are used to predict targets of phosphorylation and acetylation in multiple organisms. CAROM is influenced by the accuracy of GEMs and associated flux-balance analysis (FBA), which generate the inputs of the model. We demonstrate the protocol using multi-omics data from E. coli. For complete details on the use and execution of this protocol, please refer to Smith et al. (2022).

A flux-based machine learning model to simulate the impact of pathogen metabolic heterogeneity on drug interactions.

Drug combinations are a promising strategy to counter antibiotic resistance. However, current experimental and computational approaches do not account for the entire complexity involved in combination therapy design, such as the effect of pathogen metabolic heterogeneity, changes in the growth environment, drug treatment order, and time interval. To address these limitations, we present a comprehensive approach that uses genome-scale metabolic modeling and machine learning to guide combination therapy design. Our mechanistic approach (a) accommodates diverse data types, (b) accounts for time- and order-specific interactions, and (c) accurately predicts drug interactions in various growth conditions and their robustness to pathogen metabolic heterogeneity. Our approach achieved high accuracy (area under the receiver operating curve (AUROC) = 0.83 for synergy, AUROC = 0.98 for antagonism) in predicting drug interactions for Escherichia coli cultured in 57 metabolic conditions based on experimental validation. The entropy in bacterial metabolic response was predictive of combination therapy outcomes across time scales and growth conditions. Simulation of metabolic heterogeneity using population FBA identified two subpopulations of E. coli cells defined by the levels of three proteins (eno, fadB, and fabD) in glycolysis and lipid metabolism that influence cell tolerance to a broad range of antibiotic combinations. Analysis of the vast landscape of condition-specific drug interactions revealed a set of 24 robustly synergistic drug combinations with potential for clinical use.

Single-cell RNA-sequencing identifies anti-cancer immune phenotypes in the early lung metastatic niche during breast cancer.

Microenvironmental changes in the early metastatic niche may be exploited to identify therapeutic targets to inhibit secondary tumor formation and improve disease outcomes. We dissected the developing lung metastatic niche in a model of metastatic, triple-negative breast cancer using single-cell RNA-sequencing. Lungs were extracted from mice at 7-, 14-, or 21 days after tumor inoculation corresponding to the pre-metastatic, micro-metastatic, and metastatic niche, respectively. The progression of the metastatic niche was marked by an increase in neutrophil infiltration (5% of cells at day 0 to 81% of cells at day 21) and signaling pathways corresponding to the hallmarks of cancer. Importantly, the pre-metastatic and early metastatic niche were composed of immune cells with an anti-cancer phenotype not traditionally associated with metastatic disease. As expected, the metastatic niche exhibited pro-cancer phenotypes. The transition from anti-cancer to pro-cancer phenotypes was directly associated with neutrophil and monocyte behaviors at these time points. Predicted metabolic, transcription factor, and receptor-ligand signaling suggested that changes in the neutrophils likely induced the transitions in the other immune cells. Conditioned medium generated by cells extracted from the pre-metastatic niche successfully inhibited tumor cell proliferation and migration in vitro and the in vivo depletion of pre-metastatic neutrophils and monocytes worsened survival outcomes, thus validating the anti-cancer phenotype of the developing niche. Genes associated with the early anti-cancer response could act as biomarkers that could serve as targets for the treatment of early metastatic disease. Such therapies have the potential to revolutionize clinical outcomes in metastatic breast cancer.

Integrative analysis of clinical health records, imaging and pathogen genomics identifies personalized predictors of disease prognosis in tuberculosis.

Tuberculosis (TB) afflicts over 10 million people every year and its global burden is projected to increase dramatically due to multidrug-resistant TB (MDR-TB). The Covid-19 pandemic has resulted in reduced access to TB diagnosis and treatment, reversing decades of progress in disease management globally. It is thus crucial to analyze real-world multi-domain information from patient health records to determine personalized predictors of TB treatment outcome and drug resistance. We conduct a retrospective analysis on electronic health records of 5060 TB patients spanning 10 countries with high burden of MDR-TB including Ukraine, Moldova, Belarus and India available on the NIAID-TB portals database. We analyze over 200 features across multiple host and pathogen modalities representing patient social demographics, disease presentations as seen in cChest X rays and CT scans, and genomic records with drug susceptibility features of the pathogen strain from each patient. Our machine learning model, built with diverse data modalities outperforms models built using each modality alone in predicting treatment outcomes, with an accuracy of 81% and AUC of 0.768. We determine robust predictors across countries that are associated with unsuccessful treatmentclinical outcomes, and validate our predictions on new patient data from TB Portals. Our analysis of drug regimens and drug interactions suggests that synergistic drug combinations and those containing the drugs Bedaquiline, Levofloxacin, Clofazimine and Amoxicillin see more success in treating MDR and XDR TB. Features identified via chest imaging such as percentage of abnormal volume, size of lung cavitation and bronchial obstruction are associated significantly with pathogen genomic attributes of drug resistance. Increased disease severity was also observed in patients with lower BMI and with comorbidities. Our integrated multi-modal analysis thus revealed significant associations between radiological, microbiological, therapeutic, and demographic data modalities, providing a deeper understanding of personalized responses to aid in the clinical management of TB.

Acetyl-CoA metabolism drives epigenome change and contributes to carcinogenesis risk in fatty liver disease.

BACKGROUND: The incidence of non-alcoholic fatty liver disease (NAFLD)-associated hepatocellular carcinoma (HCC) is increasing worldwide, but the steps in precancerous hepatocytes which lead to HCC driver mutations are not well understood. Here we provide evidence that metabolically driven histone hyperacetylation in steatotic hepatocytes can increase DNA damage to initiate carcinogenesis. METHODS: Global epigenetic state was assessed in liver samples from high-fat diet or high-fructose diet rodent models, as well as in cultured immortalized human hepatocytes (IHH cells). The mechanisms linking steatosis, histone acetylation and DNA damage were investigated by computational metabolic modelling as well as through manipulation of IHH cells with metabolic and epigenetic inhibitors. Chromatin immunoprecipitation and next-generation sequencing (ChIP-seq) and transcriptome (RNA-seq) analyses were performed on IHH cells. Mutation locations and patterns were compared between the IHH cell model and genome sequence data from preneoplastic fatty liver samples from patients with alcohol-related liver disease and NAFLD. RESULTS: Genome-wide histone acetylation was increased in steatotic livers of rodents fed high-fructose or high-fat diet. In vitro, steatosis relaxed chromatin and increased DNA damage marker γH2AX, which was reversed by inhibiting acetyl-CoA production. Steatosis-associated acetylation and γH2AX were enriched at gene clusters in telomere-proximal regions which contained HCC tumour suppressors in hepatocytes and human fatty livers. Regions of metabolically driven epigenetic change also had increased levels of DNA mutation in non-cancerous tissue from NAFLD and alcohol-related liver disease patients. Finally, genome-scale network modelling indicated that redox balance could be a key contributor to this mechanism. CONCLUSIONS: Abnormal histone hyperacetylation facilitates DNA damage in steatotic hepatocytes and is a potential initiating event in hepatocellular carcinogenesis.

Lipid exposure activates gene expression changes associated with estrogen receptor negative breast cancer.

Improved understanding of local breast biology that favors the development of estrogen receptor negative (ER-) breast cancer (BC) would foster better prevention strategies. We have previously shown that overexpression of specific lipid metabolism genes is associated with the development of ER- BC. We now report results of exposure of MCF-10A and MCF-12A cells, and mammary organoids to representative medium- and long-chain polyunsaturated fatty acids. This exposure caused a dynamic and profound change in gene expression, accompanied by changes in chromatin packing density, chromatin accessibility, and histone posttranslational modifications (PTMs). We identified 38 metabolic reactions that showed significantly increased activity, including reactions related to one-carbon metabolism. Among these reactions are those that produce S-adenosyl-L-methionine for histone PTMs. Utilizing both an in-vitro model and samples from women at high risk for ER- BC, we show that lipid exposure engenders gene expression, signaling pathway activation, and histone marks associated with the development of ER- BC.

Machine learning to design antimicrobial combination therapies: Promises and pitfalls.

Combination therapies can overcome antimicrobial resistance (AMR) and repurpose existing drugs. However, the large combinatorial space to explore presents a daunting challenge. In response, machine learning (ML) algorithms are being applied to identify novel synergistic drug interactions from millions of potential combinations. Here, we compare ML-based approaches for combination therapy design based on the type of input information used, specifically: drug properties, microbial response and infection microenvironment. We also provide a compilation of publicly available drug interaction datasets relevant to AMR. Finally, we discuss limitations of current ML-based methods and propose new strategies for designing efficacious combination therapies. These include consideration of in vivo conditions, design of sequential combinations, enhancement of model interpretability and application of deep learning algorithms.

Metabolic signatures of regulation by phosphorylation and acetylation.

Acetylation and phosphorylation are highly conserved posttranslational modifications (PTMs) that regulate cellular metabolism, yet how metabolic control is shared between these PTMs is unknown. Here we analyze transcriptome, proteome, acetylome, and phosphoproteome datasets in E. coli, S. cerevisiae, and mammalian cells across diverse conditions using CAROM, a new approach that uses genome-scale metabolic networks and machine learning to classify targets of PTMs. We built a single machine learning model that predicted targets of each PTM in a condition across all three organisms based on reaction attributes (AUC>0.8). Our model predicted phosphorylated enzymes during a mammalian cell-cycle, which we validate using phosphoproteomics. Interpreting the machine learning model using game theory uncovered enzyme properties including network connectivity, essentiality, and condition-specific factors such as maximum flux that differentiate targets of phosphorylation from acetylation. The conserved and predictable partitioning of metabolic regulation identified here between these PTMs may enable rational rewiring of regulatory circuits.

The Axes of Life: A Roadmap for Understanding Dynamic Multiscale Systems.

The biological challenges facing humanity are complex, multi-factorial, and are intimately tied to the future of our health, welfare, and stewardship of the Earth. Tackling problems in diverse areas, such as agriculture, ecology, and health care require linking vast datasets that encompass numerous components and spatio-temporal scales. Here, we provide a new framework and a road map for using experiments and computation to understand dynamic biological systems that span multiple scales. We discuss theories that can help understand complex biological systems and highlight the limitations of existing methodologies and recommend data generation practices. The advent of new technologies such as big data analytics and artificial intelligence can help bridge different scales and data types. We recommend ways to make such models transparent, compatible with existing theories of biological function, and to make biological data sets readable by advanced machine learning algorithms. Overall, the barriers for tackling pressing biological challenges are not only technological, but also sociological. Hence, we also provide recommendations for promoting interdisciplinary interactions between scientists.

Deep Learning for Reintegrating Biology.

The goal of this vision paper is to investigate the possible role that advanced machine learning techniques, especially deep learning (DL), could play in the reintegration of various biological disciplines. To achieve this goal, a series of operational, but admittedly very simplistic, conceptualizations have been introduced: Life has been taken as a multidimensional phenomenon that inhabits three physical dimensions (time, space, and scale) and biological research as establishing connection between different points in the domain of life. Each of these points hence denotes a position in time, space, and scale at which a life phenomenon of interest takes place. Using these conceptualizations, fragmentation of biology can be seen as the result of too few and especially too short-ranged connections. Reintegrating biology could then be accomplished by establishing more, longer ranged connections. DL methods appear to be very well suited for addressing this particular need at this particular time. Notwithstanding the numerous unsubstantiated claims regarding the capabilities of AI, DL networks represent a major advance in the ability to find complex relationships inside large data sets that would have not been accessible with traditional data analytic methods or to a human observer. In addition, ongoing advances in the automation of taking measurements from phenomena on all levels of biological organization continue to increase the number of large quantitative data sets that are available. These increasingly common data sets could serve as anchor points for making long-range connections by virtue of DL. However, connections within the domain of life are likely to be structured in a highly nonuniform fashion and hence it is necessary to develop methods, for example, theoretical, computational, and experimental, to determine linkage of biological data sets most likely to provide useful insights on a biological problem using DL. Finally, specific DL approaches and architectures should be developed to match the needs of reintegrating biology.

Metabolism, HDACs, and HDAC Inhibitors: A Systems Biology Perspective.

Histone deacetylases (HDACs) are epigenetic enzymes that play a central role in gene regulation and are sensitive to the metabolic state of the cell. The cross talk between metabolism and histone acetylation impacts numerous biological processes including development and immune function. HDAC inhibitors are being explored for treating cancers, viral infections, inflammation, neurodegenerative diseases, and metabolic disorders. However, how HDAC inhibitors impact cellular metabolism and how metabolism influences their potency is unclear. Discussed herein are recent applications and future potential of systems biology methods such as high throughput drug screens, cancer cell line profiling, single cell sequencing, proteomics, metabolomics, and computational modeling to uncover the interplay between metabolism, HDACs, and HDAC inhibitors. The synthesis of new systems technologies can ultimately help identify epigenomic and metabolic biomarkers for patient stratification and the design of effective therapeutics.

Metabolic remodelling during early mouse embryo development.

During early mammalian embryogenesis, changes in cell growth and proliferation depend on strict genetic and metabolic instructions. However, our understanding of metabolic reprogramming and its influence on epigenetic regulation in early embryo development remains elusive. Here we show a comprehensive metabolomics profiling of key stages in mouse early development and the two-cell and blastocyst embryos, and we reconstructed the metabolic landscape through the transition from totipotency to pluripotency. Our integrated metabolomics and transcriptomics analysis shows that while two-cell embryos favour methionine, polyamine and glutathione metabolism and stay in a more reductive state, blastocyst embryos have higher metabolites related to the mitochondrial tricarboxylic acid cycle, and present a more oxidative state. Moreover, we identify a reciprocal relationship between α-ketoglutarate (α-KG) and the competitive inhibitor of α-KG-dependent dioxygenases, L-2-hydroxyglutarate (L-2-HG), where two-cell embryos inherited from oocytes and one-cell zygotes display higher L-2-HG, whereas blastocysts show higher α-KG. Lastly, increasing 2-HG availability impedes erasure of global histone methylation markers after fertilization. Together, our data demonstrate dynamic and interconnected metabolic, transcriptional and epigenetic network remodelling during early mouse embryo development.