Cross-attention enables deep learning on limited omics-imaging-clinical data of 130 lung cancer patients.

Deep-learning tools that extract prognostic factors derived from multi-omics data have recently contributed to individualized predictions of survival outcomes. However, the limited size of integrated omics-imaging-clinical datasets poses challenges. Here, we propose two biologically interpretable and robust deep-learning architectures for survival prediction of non-small cell lung cancer (NSCLC) patients, learning simultaneously from computed tomography (CT) scan images, gene expression data, and clinical information. The proposed models integrate patient-specific clinical, transcriptomic, and imaging data and incorporate Kyoto Encyclopedia of Genes and Genomes (KEGG) and Reactome pathway information, adding biological knowledge within the learning process to extract prognostic gene biomarkers and molecular pathways. While both models accurately stratify patients in high- and low-risk groups when trained on a dataset of only 130 patients, introducing a cross-attention mechanism in a sparse autoencoder significantly improves the performance, highlighting tumor regions and NSCLC-related genes as potential biomarkers and thus offering a significant methodological advancement when learning from small imaging-omics-clinical samples.

Ablation of the dystrophin Dp71f alternative C-terminal variant increases sarcoma tumour cell aggressiveness.

Alterations in Dp71 expression, the most ubiquitous dystrophin isoform, have been associated with patient survival across tumours. Intriguingly, in certain malignancies, Dp71 acts as a tumour suppressor, while manifesting oncogenic properties in others. This diversity could be explained by the expression of two Dp71 splice variants encoding proteins with distinct C-termini, each with specific properties. Expression of these variants has impeded the exploration of their unique roles. Using CRISPR/Cas9, we ablated the Dp71f variant with the alternative C-terminus in a sarcoma cell line not expressing the canonical C-terminal variant, and conducted molecular (RNAseq) and functional characterisation of the knockout cells. Dp71f ablation induced major transcriptomic alterations, particularly affecting the expression of genes involved in calcium signalling and ECM-receptor interaction pathways. The genome-scale metabolic analysis identified significant downregulation of glucose transport via membrane vesicle reaction (GLCter) and downregulated glycolysis/gluconeogenesis pathway. Functionally, these molecular changes corresponded with, increased calcium responses, cell adhesion, proliferation, survival under serum starvation and chemotherapeutic resistance. Knockout cells showed reduced GLUT1 protein expression, survival without attachment and their migration and invasion in vitro and in vivo were unaltered, despite increased matrix metalloproteinases release. Our findings emphasise the importance of alternative splicing of dystrophin transcripts and underscore the role of the Dp71f variant, which appears to govern distinct cellular processes frequently dysregulated in tumour cells. The loss of this regulatory mechanism promotes sarcoma cell survival and treatment resistance. Thus, Dp71f is a target for future investigations exploring the intricate functions of specific DMD transcripts in physiology and across malignancies.

Uncovering potential diagnostic and pathophysiological roles of α-synuclein and DJ-1 in melanoma.

BACKGROUND: Melanoma, the most lethal skin cancer type, occurs more frequently in Parkinson's disease (PD), and PD is more frequent in melanoma patients, suggesting disease mechanisms overlap. α-synuclein, a protein that accumulates in PD brain, and the oncogene DJ-1, which is associated with PD autosomal recessive forms, are both elevated in melanoma cells. Whether this indicates melanoma progression or constitutes a protective response remains unclear. We hereby investigated the molecular mechanisms through which α-synuclein and DJ-1 interact, suggesting novel biomarkers and targets in melanoma. METHODS: The Cancer Genome Atlas (TCGA) expression profiles derived from UCSC Xena were used to obtain α-synuclein and DJ-1 expression and correlated with survival in skin cutaneous melanoma (SKCM). Immunohistochemistry determined the expression in metastatic melanoma lymph nodes. Protein-protein interactions (PPIs) and molecular docking assessed protein binding and affinity with chemotherapeutic drugs. Further validation was performed using in vitro cellular models and ELISA immunoassays. RESULTS: α-synuclein and DJ-1 were upregulated in primary and metastatic SKCM. Aggregated α-synuclein was selectively detected in metastatic melanoma lymph nodes. α-synuclein overexpression in SK-MEL-28 cells induced the expression of DJ-1, supporting PPI and a positive correlation in melanoma patients. Molecular docking revealed a stable protein complex, with differential binding to chemotherapy drugs such as temozolomide, dacarbazine, and doxorubicin. Parallel reduction of both proteins in temozolomide-treated SK-MEL-28 spheroids suggests drug binding may affect protein interaction and/or stability. CONCLUSION: α-synuclein, together with DJ-1, may play a role in melanoma progression and chemosensitivity, constituting novel targets for therapeutic intervention, and possible biomarkers for melanoma.

Glycosylation spectral signatures for glioma grade discrimination using Raman spectroscopy.

BACKGROUND: Gliomas are the most common brain tumours with the high-grade glioblastoma representing the most aggressive and lethal form. Currently, there is a lack of specific glioma biomarkers that would aid tumour subtyping and minimally invasive early diagnosis. Aberrant glycosylation is an important post-translational modification in cancer and is implicated in glioma progression. Raman spectroscopy (RS), a vibrational spectroscopic label-free technique, has already shown promise in cancer diagnostics. METHODS: RS was combined with machine learning to discriminate glioma grades. Raman spectral signatures of glycosylation patterns were used in serum samples and fixed tissue biopsy samples, as well as in single cells and spheroids. RESULTS: Glioma grades in fixed tissue patient samples and serum were discriminated with high accuracy. Discrimination between higher malignant glioma grades (III and IV) was achieved with high accuracy in tissue, serum, and cellular models using single cells and spheroids. Biomolecular changes were assigned to alterations in glycosylation corroborated by analysing glycan standards and other changes such as carotenoid antioxidant content. CONCLUSION: RS combined with machine learning could pave the way for more objective and less invasive grading of glioma patients, serving as a useful tool to facilitate glioma diagnosis and delineate biomolecular glioma progression changes.

Machine Learning Methods for Survival Analysis with Clinical and Transcriptomics Data of Breast Cancer.

Breast cancer is one of the most common cancers in women worldwide, which causes an enormous number of deaths annually. However, early diagnosis of breast cancer can improve survival outcomes enabling simpler and more cost-effective treatments. The recent increase in data availability provides unprecedented opportunities to apply data-driven and machine learning methods to identify early-detection prognostic factors capable of predicting the expected survival and potential sensitivity to treatment of patients, with the final aim of enhancing clinical outcomes. This tutorial presents a protocol for applying machine learning models in survival analysis for both clinical and transcriptomic data. We show that integrating clinical and mRNA expression data is essential to explain the multiple biological processes driving cancer progression. Our results reveal that machine-learning-based models such as random survival forests, gradient boosted survival model, and survival support vector machine can outperform the traditional statistical methods, i.e., Cox proportional hazard model. The highest C-index among the machine learning models was recorded when using survival support vector machine, with a value 0.688, whereas the C-index recorded using the Cox model was 0.677. Shapley Additive Explanation (SHAP) values were also applied to identify the feature importance of the models and their impact on the prediction outcomes.

The diagnostic and prognostic potential of the EGFR/MUC4/MMP9 axis in glioma patients.

Glioblastoma is the most aggressive form of brain cancer, presenting poor prognosis despite current advances in treatment. There is therefore an urgent need for novel biomarkers and therapeutic targets. Interactions between mucin 4 (MUC4) and the epidermal growth factor receptor (EGFR) are involved in carcinogenesis, and may lead to matrix metalloproteinase-9 (MMP9) overexpression, exacerbating cancer cell invasiveness. In this study, the role of MUC4, MMP9, and EGFR in the progression and clinical outcome of glioma patients was investigated. Immunohistochemistry (IHC) and immunofluorescence (IF) in fixed tissue samples of glioma patients were used to evaluate the expression and localization of EGFR, MMP9, and MUC4. Kaplan-Meier survival analysis was also performed to test the prognostic utility of the proteins for glioma patients. The protein levels were assessed with enzyme-linked immunosorbent assay (ELISA) in serum of glioma patients, to further investigate their potential as non-invasive serum biomarkers. We demonstrated that MUC4 and MMP9 are both significantly upregulated during glioma progression. Moreover, MUC4 is co-expressed with MMP9 and EGFR in the proliferative microvasculature of glioblastoma, suggesting a potential role for MUC4 in microvascular proliferation and angiogenesis. The combined high expression of MUC4/MMP9, and MUC4/MMP9/EGFR was associated with poor overall survival (OS). Finally, MMP9 mean protein level was significantly higher in the serum of glioblastoma compared with grade III glioma patients, whereas MUC4 mean protein level was minimally elevated in higher glioma grades (III and IV) compared with control. Our results suggest that MUC4, along with MMP9, might account for glioblastoma progression, representing potential therapeutic targets, and suggesting the 'MUC4/MMP9/EGFR axis' may play a vital role in glioblastoma diagnostics.

Clinical stratification improves the diagnostic accuracy of small omics datasets within machine learning and genome-scale metabolic modelling methods.

BACKGROUND: Recently, multi-omic machine learning architectures have been proposed for the early detection of cancer. However, for rare cancers and their associated small datasets, it is still unclear how to use the available multi-omics data to achieve a mechanistic prediction of cancer onset and progression, due to the limited data available. Hepatoblastoma is the most frequent liver cancer in infancy and childhood, and whose incidence has been lately increasing in several developed countries. Even though some studies have been conducted to understand the causes of its onset and discover potential biomarkers, the role of metabolic rewiring has not been investigated in depth so far. METHODS: Here, we propose and implement an interpretable multi-omics pipeline that combines mechanistic knowledge from genome-scale metabolic models with machine learning algorithms, and we use it to characterise the underlying mechanisms controlling hepatoblastoma. RESULTS AND CONCLUSIONS: While the obtained machine learning models generally present a high diagnostic classification accuracy, our results show that the type of omics combinations used as input to the machine learning models strongly affects the detection of important genes, reactions and metabolic pathways linked to hepatoblastoma. Our method also suggests that, in the context of computer-aided diagnosis of cancer, optimal diagnostic accuracy can be achieved by adopting a combination of omics that depends on the patient's clinical characteristics.

A Practical Guide to Integrating Multimodal Machine Learning and Metabolic Modeling.

Complex, distributed, and dynamic sets of clinical biomedical data are collectively referred to as multimodal clinical data. In order to accommodate the volume and heterogeneity of such diverse data types and aid in their interpretation when they are combined with a multi-scale predictive model, machine learning is a useful tool that can be wielded to deconstruct biological complexity and extract relevant outputs. Additionally, genome-scale metabolic models (GSMMs) are one of the main frameworks striving to bridge the gap between genotype and phenotype by incorporating prior biological knowledge into mechanistic models. Consequently, the utilization of GSMMs as a foundation for the integration of multi-omic data originating from different domains is a valuable pursuit towards refining predictions. In this chapter, we show how cancer multi-omic data can be analyzed via multimodal machine learning and metabolic modeling. Firstly, we focus on the merits of adopting an integrative systems biology led approach to biomedical data mining. Following this, we propose how constraint-based metabolic models can provide a stable yet adaptable foundation for the integration of multimodal data with machine learning. Finally, we provide a step-by-step tutorial for the combination of machine learning and GSMMs, which includes: (i) tissue-specific constraint-based modeling; (ii) survival analysis using time-to-event prediction for cancer; and (iii) classification and regression approaches for multimodal machine learning. The code associated with the tutorial can be found at https://github.com/Angione-Lab/Tutorials_Combining_ML_and_GSMM .

Computational profiling of natural compounds as promising inhibitors against the spike proteins of SARS-CoV-2 wild-type and the variants of concern, viral cell-entry process, and cytokine storm in COVID-19.

The continuous spread and evolution of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and the rapid surge in infection cases in the coronavirus disease 2019 (COVID-19) evoke a dire need for effective therapeutics. In this study, we explored the inhibitory potential of a library of 605 phytocompounds, selected from Indian medicinal plants with reported antiviral and anti-inflammatory activities, against the receptor-binding domain of spike proteins of the SARS-CoV-2 wild-type and the variants of concern, including variants B.1.1.7 (Alpha), B.1.351 (Beta), P.1 (Gamma), B.1.617.2 (Delta), and B.1.1.529 (Omicron). Our approach was based on extensive molecular docking, assessment of drug-likeness, and robust molecular dynamics simulations. We also identified promising inhibitory candidates against the host (human) proteins associated with SARS-CoV-2 spike activation and attachment, namely, ACE2 receptor, proteases TMPRSS2 and CTSL, and the endocytic regulator AAK1. In addition, we screened promising inhibitory compounds against the human proinflammatory cytokines- IL-6, IL-1ß, TNF-α, and IFN-γ, that are associated with the adverse cytokine storm in COVID-19 patients. Our analysis returned an encouraging list of promising inhibitory candidates that includes: abietatriene against the spike proteins of the SARS-CoV-2 wild-type and the variants of concern; taraxerol against the human ACE2, CTSL and TNF-α; ß-amyrin against the human TMPRSS2; cynaroside against the human AAK1 and IL-1ß; and friedelin against the human IL-6 and IFN-γ. Our findings provide substantial evidence for the inhibitory potential of these compounds and encourage further in vitro and in vivo studies to validate their use as safe and effective therapeutics against COVID-19.

Genome-scale metabolic modelling of SARS-CoV-2 in cancer cells reveals an increased shift to glycolytic energy production.

Cancer is considered a high-risk condition for severe illness resulting from COVID-19. The interaction between severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) and human metabolism is key to elucidating the risk posed by COVID-19 for cancer patients and identifying effective treatments, yet it is largely uncharacterised on a mechanistic level. We present a genome-scale map of short-term metabolic alterations triggered by SARS-CoV-2 infection of cancer cells. Through transcriptomic- and proteomic-informed genome-scale metabolic modelling, we characterise the role of RNA and fatty acid biosynthesis in conjunction with a rewiring in energy production pathways and enhanced cytokine secretion. These findings link together complementary aspects of viral invasion of cancer cells, while providing mechanistic insights that can inform the development of treatment strategies.

Integrated multi-omics analysis of ovarian cancer using variational autoencoders.

Cancer is a complex disease that deregulates cellular functions at various molecular levels (e.g., DNA, RNA, and proteins). Integrated multi-omics analysis of data from these levels is necessary to understand the aberrant cellular functions accountable for cancer and its development. In recent years, Deep Learning (DL) approaches have become a useful tool in integrated multi-omics analysis of cancer data. However, high dimensional multi-omics data are generally imbalanced with too many molecular features and relatively few patient samples. This imbalance makes a DL based integrated multi-omics analysis difficult. DL-based dimensionality reduction technique, including variational autoencoder (VAE), is a potential solution to balance high dimensional multi-omics data. However, there are few VAE-based integrated multi-omics analyses, and they are limited to pancancer. In this work, we did an integrated multi-omics analysis of ovarian cancer using the compressed features learned through VAE and an improved version of VAE, namely Maximum Mean Discrepancy VAE (MMD-VAE). First, we designed and developed a DL architecture for VAE and MMD-VAE. Then we used the architecture for mono-omics, integrated di-omics and tri-omics data analysis of ovarian cancer through cancer samples identification, molecular subtypes clustering and classification, and survival analysis. The results show that MMD-VAE and VAE-based compressed features can respectively classify the transcriptional subtypes of the TCGA datasets with an accuracy in the range of 93.2-95.5% and 87.1-95.7%. Also, survival analysis results show that VAE and MMD-VAE based compressed representation of omics data can be used in cancer prognosis. Based on the results, we can conclude that (i) VAE and MMD-VAE outperform existing dimensionality reduction techniques, (ii) integrated multi-omics analyses perform better or similar compared to their mono-omics counterparts, and (iii) MMD-VAE performs better than VAE in most omics dataset.

Discovering Essential Multiple Gene Effects Through Large Scale Optimization: An Application to Human Cancer Metabolism.

Computational modelling of metabolic processes has proven to be a useful approach to formulate our knowledge and improve our understanding of core biochemical systems that are crucial to maintaining cellular functions. Towards understanding the broader role of metabolism on cellular decision-making in health and disease conditions, it is important to integrate the study of metabolism with other core regulatory systems and omics within the cell, including gene expression patterns. After quantitatively integrating gene expression profiles with a genome-scale reconstruction of human metabolism, we propose a set of combinatorial methods to reverse engineer gene expression profiles and to find pairs and higher-order combinations of genetic modifications that simultaneously optimize multi-objective cellular goals. This enables us to suggest classes of transcriptomic profiles that are most suitable to achieve given metabolic phenotypes. We demonstrate how our techniques are able to compute beneficial, neutral or "toxic" combinations of gene expression levels. We test our methods on nine tissue-specific cancer models, comparing our outcomes with the corresponding normal cells, identifying genes as targets for potential therapies. Our methods open the way to a broad class of applications that require an understanding of the interplay among genotype, metabolism, and cellular behaviour, at scale.

Integrating splice-isoform expression into genome-scale models characterizes breast cancer metabolism.

Motivation: Despite being often perceived as the main contributors to cell fate and physiology, genes alone cannot predict cellular phenotype. During the process of gene expression, 95% of human genes can code for multiple proteins due to alternative splicing. While most splice variants of a gene carry the same function, variants within some key genes can have remarkably different roles. To bridge the gap between genotype and phenotype, condition- and tissue-specific models of metabolism have been constructed. However, current metabolic models only include information at the gene level. Consequently, as recently acknowledged by the scientific community, common situations where changes in splice-isoform expression levels alter the metabolic outcome cannot be modeled. Results: We here propose GEMsplice, the first method for the incorporation of splice-isoform expression data into genome-scale metabolic models. Using GEMsplice, we make full use of RNA-Seq quantitative expression profiles to predict, for the first time, the effects of splice isoform-level changes in the metabolism of 1455 patients with 31 different breast cancer types. We validate GEMsplice by generating cancer-versus-normal predictions on metabolic pathways, and by comparing with gene-level approaches and available literature on pathways affected by breast cancer. GEMsplice is freely available for academic use at https://github.com/GEMsplice/GEMsplice_code. Compared to state-of-the-art methods, we anticipate that GEMsplice will enable for the first time computational analyses at transcript level with splice-isoform resolution. Availability and implementation: https://github.com/GEMsplice/GEMsplice_code. Contact: c.angione@tees.ac.uk. Supplementary information: Supplementary data are available at Bioinformatics online.

Modelling pyruvate dehydrogenase under hypoxia and its role in cancer metabolism.

Metabolism is the only biological system that can be fully modelled at genome scale. As a result, metabolic models have been increasingly used to study the molecular mechanisms of various diseases. Hypoxia, a low-oxygen tension, is a well-known characteristic of many cancer cells. Pyruvate dehydrogenase (PDH) controls the flux of metabolites between glycolysis and the tricarboxylic acid cycle and is a key enzyme in metabolic reprogramming in cancer metabolism. Here, we develop and manually curate a constraint-based metabolic model to investigate the mechanism of pyruvate dehydrogenase under hypoxia. Our results characterize the activity of pyruvate dehydrogenase and its decline during hypoxia. This results in lactate accumulation, consistent with recent hypoxia studies and a well-known feature in cancer metabolism. We apply machine-learning techniques on the flux datasets to identify reactions that drive these variations. We also identify distinct features on the structure of the variables and individual metabolic components in the switch from normoxia to hypoxia. Our results provide a framework for future studies by integrating multi-omics data to predict condition-specific metabolic phenotypes under hypoxia.

Multi-Target Analysis and Design of Mitochondrial Metabolism.

Analyzing and optimizing biological models is often identified as a research priority in biomedical engineering. An important feature of a model should be the ability to find the best condition in which an organism has to be grown in order to reach specific optimal output values chosen by the researcher. In this work, we take into account a mitochondrial model analyzed with flux-balance analysis. The optimal design and assessment of these models is achieved through single- and/or multi-objective optimization techniques driven by epsilon-dominance and identifiability analysis. Our optimization algorithm searches for the values of the flux rates that optimize multiple cellular functions simultaneously. The optimization of the fluxes of the metabolic network includes not only input fluxes, but also internal fluxes. A faster convergence process with robust candidate solutions is permitted by a relaxed Pareto dominance, regulating the granularity of the approximation of the desired Pareto front. We find that the maximum ATP production is linked to a total consumption of NADH, and reaching the maximum amount of NADH leads to an increasing request of NADH from the external environment. Furthermore, the identifiability analysis characterizes the type and the stage of three monogenic diseases. Finally, we propose a new methodology to extend any constraint-based model using protein abundances.

Pareto optimality in organelle energy metabolism analysis.

In low and high eukaryotes, energy is collected or transformed in compartments, the organelles. The rich variety of size, characteristics, and density of the organelles makes it difficult to build a general picture. In this paper, we make use of the Pareto-front analysis to investigate the optimization of energy metabolism in mitochondria and chloroplasts. Using the Pareto optimality principle, we compare models of organelle metabolism on the basis of single- and multiobjective optimization, approximation techniques (the Bayesian Automatic Relevance Determination), robustness, and pathway sensitivity analysis. Finally, we report the first analysis of the metabolic model for the hydrogenosome of Trichomonas vaginalis, which is found in several protozoan parasites. Our analysis has shown the importance of the Pareto optimality for such comparison and for insights into the evolution of the metabolism from cytoplasmic to organelle bound, involving a model order reduction. We report that Pareto fronts represent an asymptotic analysis useful to describe the metabolism of an organism aimed at maximizing concurrently two or more metabolite concentrations.