Advancing drug-response prediction using multi-modal and -omics machine learning integration (MOMLIN): a case study on breast cancer clinical data.

The inherent heterogeneity of cancer contributes to highly variable responses to any anticancer treatments. This underscores the need to first identify precise biomarkers through complex multi-omics datasets that are now available. Although much research has focused on this aspect, identifying biomarkers associated with distinct drug responders still remains a major challenge. Here, we develop MOMLIN, a multi-modal and -omics machine learning integration framework, to enhance drug-response prediction. MOMLIN jointly utilizes sparse correlation algorithms and class-specific feature selection algorithms, which identifies multi-modal and -omics-associated interpretable components. MOMLIN was applied to 147 patients' breast cancer datasets (clinical, mutation, gene expression, tumor microenvironment cells and molecular pathways) to analyze drug-response class predictions for non-responders and variable responders. Notably, MOMLIN achieves an average AUC of 0.989, which is at least 10% greater when compared with current state-of-the-art (data integration analysis for biomarker discovery using latent components, multi-omics factor analysis, sparse canonical correlation analysis). Moreover, MOMLIN not only detects known individual biomarkers such as genes at mutation/expression level, most importantly, it correlates multi-modal and -omics network biomarkers for each response class. For example, an interaction between ER-negative-HMCN1-COL5A1 mutations-FBXO2-CSF3R expression-CD8 emerge as a multimodal biomarker for responders, potentially affecting antimicrobial peptides and FLT3 signaling pathways. In contrast, for resistance cases, a distinct combination of lymph node-TP53 mutation-PON3-ENSG00000261116 lncRNA expression-HLA-E-T-cell exclusions emerged as multimodal biomarkers, possibly impacting neurotransmitter release cycle pathway. MOMLIN, therefore, is expected advance precision medicine, such as to detect context-specific multi-omics network biomarkers and better predict drug-response classifications.

Towards multi-omics synthetic data integration.

Across many scientific disciplines, the development of computational models and algorithms for generating artificial or synthetic data is gaining momentum. In biology, there is a great opportunity to explore this further as more and more big data at multi-omics level are generated recently. In this opinion, we discuss the latest trends in biological applications based on process-driven and data-driven aspects. Moving ahead, we believe these methodologies can help shape novel multi-omics-scale cellular inferences.

Machine learning alternative to systems biology should not solely depend on data.

In recent years, artificial intelligence (AI)/machine learning has emerged as a plausible alternative to systems biology for the elucidation of biological phenomena and in attaining specified design objective in synthetic biology. Although considered highly disruptive with numerous notable successes so far, we seek to bring attention to both the fundamental and practical pitfalls of their usage, especially in illuminating emergent behaviors from chaotic or stochastic systems in biology. Without deliberating on their suitability and the required data qualities and pre-processing approaches beforehand, the research and development community could experience similar 'AI winters' that had plagued other fields. Instead, we anticipate the integration or combination of the two approaches, where appropriate, moving forward.

Identifying toggle genes from transcriptome-wide scatter: A new perspective for biological regulation.

The study of gene expression variability, especially for cancer and cell differentiation studies, has become important. Here, we investigate transcriptome-wide scatter of 23 cell types and conditions across different levels of biological complexity. We focused on genes that act like toggle switches between pairwise replicates of the same cell type, i.e. genes expressed in one replicate and not expressed in the other, sometimes also referred as ON/OFF genes. The proportion of these toggle genes dramatically increases from unicellular to multicellular organization, especially for development and cancer cells. A relevant portion of toggle switches are non-coding genes: in unicellular systems the most represented classes are tRNA and rRNA, while multicellular systems more frequently show lncRNA, sncRNA and pseudogenes. Notably, disease associated microRNAs (miRNAs), pseudogenes and numerous uncharacterized transcripts are present in both development and cancer cells. On top of the known intrinsic and extrinsic factors, our work indicates toggle genes as a novel collective component creating transcriptome-wide variability. This requires further investigation for elucidating both evolutionary and disease processes.

The reduction of gene expression variability from single cells to populations follows simple statistical laws.

Recent studies on single cells and population transcriptomics have revealed striking differences in global gene expression distributions. Single cells display highly variable expressions between cells, while cell populations present deterministic global patterns. The mechanisms governing the reduction of transcriptome-wide variability over cell ensemble size, however, remain largely unknown. To investigate transcriptome-wide variability of single cells to different sizes of cell populations, we examined RNA-Seq datasets of 6 mammalian cell types. Our statistical analyses show, for each cell type, increasing cell ensemble size reduces scatter in transcriptome-wide expressions and noise (variance over square mean) values, with corresponding increases in Pearson and Spearman correlations. Next, accounting for technical variability by the removal of lowly expressed transcripts, we demonstrate that transcriptome-wide variability reduces, approximating the law of large numbers. Subsequent analyses reveal that the entire gene expressions of cell populations and only the highly expressed portion of single cells are Gaussian distributed, following the central limit theorem.

Identifying Key In Silico Knockout for Enhancement of Limonene Yield Through Dynamic Metabolic Modelling.

Living cells display dynamic and complex behaviors. To understand their response and to infer novel insights not possible with traditional reductionist approaches, over the last few decades various computational modelling methodologies have been developed. In this chapter, we focus on modelling the dynamic metabolic response, using linear and nonlinear ordinary differential equations, of an engineered Escherichia coli MG1655 strain with plasmid pJBEI-6409 that produces limonene. We show the systems biology steps involved from collecting time-series data of living cells, to dynamic model creation and fitting the model with experimental responses using COPASI software.

Can digital twin efforts shape microorganism-based alternative food?

With the continuous increment in global population growth, compounded by post-pandemic food security challenges due to labor shortages, effects of climate change, political conflicts, limited land for agriculture, and carbon emissions control, addressing food production in a sustainable manner for future generations is critical. Microorganisms are potential alternative food sources that can help close the gap in food production. For the development of more efficient and yield-enhancing products, it is necessary to have a better understanding on the underlying regulatory molecular pathways of microbial growth. Nevertheless, as microbes are regulated at multiomics scales, current research focusing on single omics (genomics, proteomics, or metabolomics) independently is inadequate for optimizing growth and product output. Here, we discuss digital twin (DT) approaches that integrate systems biology and artificial intelligence in analyzing multiomics datasets to yield a microbial replica model for in silico testing before production. DT models can thus provide a holistic understanding of microbial growth, metabolite biosynthesis mechanisms, as well as identifying crucial production bottlenecks. Our argument, therefore, is to support the development of novel DT models that can potentially revolutionize microorganism-based alternative food production efficiency.

A concerted increase in readthrough and intron retention drives transposon expression during aging and senescence.

Aging and senescence are characterized by pervasive transcriptional dysfunction, including increased expression of transposons and introns. Our aim was to elucidate mechanisms behind this increased expression. Most transposons are found within genes and introns, with a large minority being close to genes. This raises the possibility that transcriptional readthrough and intron retention are responsible for age-related changes in transposon expression rather than expression of autonomous transposons. To test this, we compiled public RNA-seq datasets from aged human fibroblasts, replicative and drug-induced senescence in human cells, and RNA-seq from aging mice and senescent mouse cells. Indeed, our reanalysis revealed a correlation between transposons expression, intron retention, and transcriptional readthrough across samples and within samples. Both intron retention and readthrough increased with aging or cellular senescence and these transcriptional defects were more pronounced in human samples as compared to those of mice. In support of a causal connection between readthrough and transposon expression, analysis of models showing induced transcriptional readthrough confirmed that they also show elevated transposon expression. Taken together, our data suggest that elevated transposon reads during aging seen in various RNA-seq dataset are concomitant with multiple transcriptional defects. Intron retention and transcriptional readthrough are the most likely explanation for the expression of transposable elements that lack a functional promoter.

A systems biology approach to suppress TNF-induced proinflammatory gene expressions.

BACKGROUND: Tumor necrosis factor (TNF) is a widely studied cytokine (ligand) that induces proinflammatory signaling and regulates myriad cellular processes. In major illnesses, such as rheumatoid arthritis and certain cancers, the expression of TNF is elevated. Despite much progress in the field, the targeted regulation of TNF response for therapeutic benefits remains suboptimal. Here, to effectively regulate the proinflammatory response induced by TNF, a systems biology approach was adopted. RESULTS: We developed a computational model to investigate the temporal activations of MAP kinase (p38), nuclear factor (NF)-κB, and the kinetics of 3 groups of genes, defined by early, intermediate and late phases, in murine embryonic fibroblast (MEF) and 3T3 cells. To identify a crucial target that suppresses, and not abolishes, proinflammatory genes, the model was tested in several in silico knock out (KO) conditions. Among the candidate molecules tested, in silico RIP1 KO effectively regulated all groups of proinflammatory genes (early, middle and late). To validate this result, we experimentally inhibited TNF signaling in MEF and 3T3 cells with RIP1 inhibitor, Necrostatin-1 (Nec-1), and investigated 10 genes (Il6, Nfkbia, Jun, Tnfaip3, Ccl7, Vcam1, Cxcl10, Mmp3, Mmp13, Enpp2) belonging to the 3 major groups of upregulated genes. As predicted by the model, all measured genes were significantly impaired. CONCLUSIONS: Our results demonstrate that Nec-1 modulates TNF-induced proinflammatory response, and may potentially be used as a therapeutic target for inflammatory diseases such as rheumatoid arthritis and osteoarthritis.

Systems Biology and Omics Approaches for Complex Human Diseases.

For many years, there has been general interest in developing virtual cells or digital twin models [...].

Application of GeneCloudOmics: Transcriptomic Data Analytics for Synthetic Biology.

Research in synthetic biology and metabolic engineering require a deep understanding on the function and regulation of complex pathway genes. This can be achieved through gene expression profiling which quantifies the transcriptome-wide expression under any condition, such as a cell development stage, mutant, disease, or treatment with a drug. The expression profiling is usually done using high-throughput techniques such as RNA sequencing (RNA-Seq) or microarray. Although both methods are based on different technical approaches, they provide quantitative measures of the expression levels of thousands of genes. The expression levels of the genes are compared under different conditions to identify the differentially expressed genes (DEGs), the genes with different expression levels under different conditions. DEGs, usually involving thousands in number, are then investigated using bioinformatics and data analytic tools to infer and compare their functional roles between conditions. Dealing with such large datasets, therefore, requires intensive data processing and analyses to ensure its quality and produce results that are statistically sound. Thus, there is a need for deep statistical and bioinformatics knowledge to deal with high-throughput gene expression data. This represents a barrier for wet biologists with limited computational, programming, and data analytic skills that prevent them from getting the full potential of the data. In this chapter, we present a step-by-step protocol to perform transcriptome analysis using GeneCloudOmics, a cloud-based web server that provides an end-to-end platform for high-throughput gene expression analysis.

Globally invariant behavior of oncogenes and random genes at population but not at single cell level.

Cancer is widely considered a genetic disease. Notably, recent works have highlighted that every human gene may possibly be associated with cancer. Thus, the distinction between genes that drive oncogenesis and those that are associated to the disease, but do not play a role, requires attention. Here we investigated single cells and bulk (cell-population) datasets of several cancer transcriptomes and proteomes in relation to their healthy counterparts. When analyzed by machine learning and statistical approaches in bulk datasets, both general and cancer-specific oncogenes, as defined by the Cancer Genes Census, show invariant behavior to randomly selected gene sets of the same size for all cancers. However, when protein-protein interaction analyses were performed, the oncogenes-derived networks show higher connectivity than those relative to random genes. Moreover, at single-cell scale, we observe variant behavior in a subset of oncogenes for each considered cancer type. Moving forward, we concur that the role of oncogenes needs to be further scrutinized by adopting protein causality and higher-resolution single-cell analyses.

Perspective: Multiomics and Machine Learning Help Unleash the Alternative Food Potential of Microalgae.

Food security has become a pressing issue in the modern world. The ever-increasing world population, ongoing COVID-19 pandemic, and political conflicts together with climate change issues make the problem very challenging. Therefore, fundamental changes to the current food system and new sources of alternative food are required. Recently, the exploration of alternative food sources has been supported by numerous governmental and research organizations, as well as by small and large commercial ventures. Microalgae are gaining momentum as an effective source of alternative laboratory-based nutritional proteins as they are easy to grow under variable environmental conditions, with the added advantage of absorbing carbon dioxide. Despite their attractiveness, the utilization of microalgae faces several practical limitations. Here, we discuss both the potential and challenges of microalgae in food sustainability and their possible long-term contribution to the circular economy of converting food waste into feed via modern methods. We also argue that systems biology and artificial intelligence can play a role in overcoming some of the challenges and limitations; through data-guided metabolic flux optimization, and by systematically increasing the growth of the microalgae strains without negative outcomes, such as toxicity. This requires microalgae databases rich in omics data and further developments on its mining and analytics methods.

Metabolomics and modelling approaches for systems metabolic engineering.

Metabolic engineering involves the manipulation of microbes to produce desirable compounds through genetic engineering or synthetic biology approaches. Metabolomics involves the quantitation of intracellular and extracellular metabolites, where mass spectrometry and nuclear magnetic resonance based analytical instrumentation are often used. Here, the experimental designs, sample preparations, metabolite quenching and extraction are essential to the quantitative metabolomics workflow. The resultant metabolomics data can then be used with computational modelling approaches, such as kinetic and constraint-based modelling, to better understand underlying mechanisms and bottlenecks in the synthesis of desired compounds, thereby accelerating research through systems metabolic engineering. Constraint-based models, such as genome scale models, have been used successfully to enhance the yield of desired compounds from engineered microbes, however, unlike kinetic or dynamic models, constraint-based models do not incorporate regulatory effects. Nevertheless, the lack of time-series metabolomic data generation has hindered the usefulness of dynamic models till today. In this review, we show that improvements in automation, dynamic real-time analysis and high throughput workflows can drive the generation of more quality data for dynamic models through time-series metabolomics data generation. Spatial metabolomics also has the potential to be used as a complementary approach to conventional metabolomics, as it provides information on the localization of metabolites. However, more effort must be undertaken to identify metabolites from spatial metabolomics data derived through imaging mass spectrometry, where machine learning approaches could prove useful. On the other hand, single-cell metabolomics has also seen rapid growth, where understanding cell-cell heterogeneity can provide more insights into efficient metabolic engineering of microbes. Moving forward, with potential improvements in automation, dynamic real-time analysis, high throughput workflows, and spatial metabolomics, more data can be produced and studied using machine learning algorithms, in conjunction with dynamic models, to generate qualitative and quantitative predictions to advance metabolic engineering efforts.

Macroscopic law of conservation revealed in the population dynamics of Toll-like receptor signaling.

Stimulating the receptors of a single cell generates stochastic intracellular signaling. The fluctuating response has been attributed to the low abundance of signaling molecules and the spatio-temporal effects of diffusion and crowding. At population level, however, cells are able to execute well-defined deterministic biological processes such as growth, division, differentiation and immune response. These data reflect biology as a system possessing microscopic and macroscopic dynamics. This commentary discusses the average population response of the Toll-like receptor (TLR) 3 and 4 signaling. Without requiring detailed experimental data, linear response equations together with the fundamental law of information conservation have been used to decipher novel network features such as unknown intermediates, processes and cross-talk mechanisms. For single cell response, however, such simplicity seems far from reality. Thus, as observed in any other complex systems, biology can be considered to possess order and disorder, inheriting a mixture of predictable population level and unpredictable single cell outcomes.

Systems Biology to Understand and Regulate Human Retroviral Proinflammatory Response.

The majority of human genome are non-coding genes. Recent research have revealed that about half of these genome sequences make up of transposable elements (TEs). A branch of these belong to the endogenous retroviruses (ERVs), which are germline viral infection that occurred over millions of years ago. They are generally harmless as evolutionary mutations have made them unable to produce viral agents and are mostly epigenetically silenced. Nevertheless, ERVs are able to express by still unknown mechanisms and recent evidences have shown links between ERVs and major proinflammatory diseases and cancers. The major challenge is to elucidate a detailed mechanistic understanding between them, so that novel therapeutic approaches can be explored. Here, we provide a brief overview of TEs, human ERVs and their links to microbiome, innate immune response, proinflammatory diseases and cancer. Finally, we recommend the employment of systems biology approaches for future HERV research.

GeneCloudOmics: A Data Analytic Cloud Platform for High-Throughput Gene Expression Analysis.

Gene expression profiling techniques, such as DNA microarray and RNA-Sequencing, have provided significant impact on our understanding of biological systems. They contribute to almost all aspects of biomedical research, including studying developmental biology, host-parasite relationships, disease progression and drug effects. However, the high-throughput data generations present challenges for many wet experimentalists to analyze and take full advantage of such rich and complex data. Here we present GeneCloudOmics, an easy-to-use web server for high-throughput gene expression analysis that extends the functionality of our previous ABioTrans with several new tools, including protein datasets analysis, and a web interface. GeneCloudOmics allows both microarray and RNA-Seq data analysis with a comprehensive range of data analytics tools in one package that no other current standalone software or web-based tool can do. In total, GeneCloudOmics provides the user access to 23 different data analytical and bioinformatics tasks including reads normalization, scatter plots, linear/non-linear correlations, PCA, clustering (hierarchical, k-means, t-SNE, SOM), differential expression analyses, pathway enrichments, evolutionary analyses, pathological analyses, and protein-protein interaction (PPI) identifications. Furthermore, GeneCloudOmics allows the direct import of gene expression data from the NCBI Gene Expression Omnibus database. The user can perform all tasks rapidly through an intuitive graphical user interface that overcomes the hassle of coding, installing tools/packages/libraries and dealing with operating systems compatibility and version issues, complications that make data analysis tasks challenging for biologists. Thus, GeneCloudOmics is a one-stop open-source tool for gene expression data analysis and visualization. It is freely available at http://combio-sifbi.org/GeneCloudOmics.

Attractor Concepts to Evaluate the Transcriptome-wide Dynamics Guiding Anaerobic to Aerobic State Transition in Escherichia coli.

For any dynamical system, like living organisms, an attractor state is a set of variables or mechanisms that converge towards a stable system behavior despite a wide variety of initial conditions. Here, using multi-dimensional statistics, we investigate the global gene expression attractor mechanisms shaping anaerobic to aerobic state transition (AAT) of Escherichia coli in a bioreactor at early times. Out of 3,389 RNA-Seq expression changes over time, we identified 100 sharply changing genes that are key for guiding 1700 genes into the AAT attractor basin. Collectively, these genes were named as attractor genes constituting of 6 dynamic clusters. Apart from the expected anaerobic (glycolysis), aerobic (TCA cycle) and fermentation (succinate pathways) processes, sulphur metabolism, ribosome assembly and amino acid transport mechanisms together with 332 uncharacterised genes are also key for AAT. Overall, our work highlights the importance of multi-dimensional statistical analyses for revealing novel processes shaping AAT.

ScatLay: utilizing transcriptome-wide noise for identifying and visualizing differentially expressed genes.

Differential expressed (DE) genes analysis is valuable for understanding comparative transcriptomics between cells, conditions or time evolution. However, the predominant way of identifying DE genes is to use arbitrary threshold fold or expression changes as cutoff. Here, we developed a more objective method, Scatter Overlay or ScatLay, to extract and graphically visualize DE genes across any two samples by utilizing their pair-wise scatter or transcriptome-wide noise, while factoring replicate variabilities. We tested ScatLay for 3 cell types: between time points for Escherichia coli aerobiosis and Saccharomyces cerevisiae hypoxia, and between untreated and Etomoxir treated Mus Musculus embryonic stem cell. As a result, we obtain 1194, 2061 and 2932 DE genes, respectively. Next, we compared these data with two widely used current approaches (DESeq2 and NOISeq) with typical twofold expression changes threshold, and show that ScatLay reveals significantly larger number of DE genes. Hence, our method provides a wider coverage of DE genes, and will likely pave way for finding more novel regulatory genes in future works.

Systems biology approaches integrated with artificial intelligence for optimized metabolic engineering.

Metabolic engineering aims to maximize the production of bio-economically important substances (compounds, enzymes, or other proteins) through the optimization of the genetics, cellular processes and growth conditions of microorganisms. This requires detailed understanding of underlying metabolic pathways involved in the production of the targeted substances, and how the cellular processes or growth conditions are regulated by the engineering. To achieve this goal, a large system of experimental techniques, compound libraries, computational methods and data resources, including multi-omics data, are used. The recent advent of multi-omics systems biology approaches significantly impacted the field by opening new avenues to perform dynamic and large-scale analyses that deepen our knowledge on the manipulations. However, with the enormous transcriptomics, proteomics and metabolomics available, it is a daunting task to integrate the data for a more holistic understanding. Novel data mining and analytics approaches, including Artificial Intelligence (AI), can provide breakthroughs where traditional low-throughput experiment-alone methods cannot easily achieve. Here, we review the latest attempts of combining systems biology and AI in metabolic engineering research, and highlight how this alliance can help overcome the current challenges facing industrial biotechnology, especially for food-related substances and compounds using microorganisms.