Enhancing Omics Analyses Through Coalitional Games and Shapley Values.

OPINION article Front. Physiol., 06 November 2015Sec. Systems Biology Archive Volume 6 - 2015 | https://doi.org/10.3389/fphys.2015.00318

The coexistence of LTE-Unlicensed (LTE-U) and WiFi in unlicensed spectrum is studied in the context of airtime sharing. We consider core problem where a set of LTE-U cells from different operators share the same channel as a co-located WiFi access point (AP). We assume that LTE-U cells utilize Listen-Before-Talk (LBT) as the default channel access mechanism. Principally, we deal with the following question: how should an operator's LTE-U cell adjust its contention window in order to provide a fair coexistence both with WiFi and co-located LTE-U cells of other operators? We consider that LTE-U cells behave altruistically both among themselves and to WiFi. Cooperation of LTE-U cells is studied using a coalition formation game framework which is based on the well-known Shapley value. We define a payoff configuration scheme in the coalition game which involves altruism. We prove that the coalitional game is always zero-monotonic, and Shapley value is also max-min fair. We compare airtime sharing performance of Shapley value with weighted proportional fairness via numerical results and show that Shapley value provides much better fairness than proportional fairness as determined by entropy and Jain's index metrics while having roughly equal average airtime.

The past half century has seen a tremendous progress in technology. The first hand held mobile phones began production in 1973[1] and today about 45% of the world's population owns a smart phone.[2] In 1983, the publicly available and practical version of the internet was invented allowing standard computer networks to communicate with one another.[3] Today, it is rare to find a stand-alone computer not connected to the world wide web. Science took a historical leap forward approximately a quarter of a century ago with the development of the human genome project (HGP).[4] While the HGP was the complete mapping of all the genes in the human genome, ultimately it changed how science was approached. The ushering in of genomics via the HGP quickly led to en masse measurement of gene transcripts (transcriptomics), proteins (proteomics), and finally metabolites (metabolomics). While the knowledge gained from these branches of omics will continue to increase as time passes, one thing we have learned for certain is that the human cell is an amazing feat of engineering. Its genome is made up of approximately 20 000 protein coding genes and produces an estimated 200 000 transcripts. Within the genomic code are blueprints for potentially producing upwards of a billion protein species when various splice variants and post-translational modifications are considered. The human metabolite database currently lists over 114 000 metabolites that have either been detected or are predicted to be present within human cells.[5] Compare these numbers to an automobile, which contains on the order of tens of thousands of moving parts. Just like an automobile engine, the components of the cell work together to create a functional entity. Also like an engine, the source of a dysfunctional component within a cell may not be directly connected to the component itself. For example, rapid cell proliferation in human epidermal growth factor receptor 2 (HER-2) breast cancer is caused by an excess of HER-2 receptors on the surface of the cells.[6] The source of this receptor expansion, however, can be traced to gene duplication events. Historically, studies would be designed to identify this connection molecule by molecule, specifically targeting the protein receptor and HER-2 gene. In today's omics-era it is possible to collect data on thousands of genes, transcripts, and proteins so that not only the source of the biological problem can be detected but also the peripheral damage it creates. The systems biology era has given rise to a plethora of different omics fields that aim to characterize various aspects of cells and organisms. The foundation of systems biology, however, remains the "big-four" omics: genomics, transcriptomics, proteomics, and metabolomics. The initial article in this edition provides a reflection on the progress that has been made in omics research beginning with the HGP, which arguably changed the strategy in how scientists design biological studies.[7] The article highlights the incredible progress that has been made in the big-four omic fields. Description of an article detailing a multi-omics study of plasma samples obtained from individuals infected with COVID-19 demonstrates how mainstream systems biology studies have become in just the past few years. While the article also highlights challenges in systems biology, it provides a glimpse into the many different databases and consortiums that have arisen to address the ambitious nature of attempting to characterize all the components within cells and organisms. While much effort has been spent in the past couple of decades identifying the entire complement of cellular biomolecules, the need to identify how these biomolecules interact to carry out the many functions that occur within these cells has become paramount. In their article, Basu et al., describe the use of functional proteomics and systems biology to generate protein-protein interaction networks in neurons, glia, and other cell types.[8] They summarize findings from recent studies that compare these interaction networks in normally functioning neurons, in developing neurons, and neurodegenerative diseases. The article goes on to demonstrate how these interaction networks can be used to decipher cellular functions that regulate neuronal activity and brain function. While the era of systems biology has increased the amount of data per study, most laboratories are not historically designed to accommodate this requirement. A complete systems biology view of a cell requires data from a variety of classes of biomolecules (e.g., DNA, RNA, proteins, etc.), however, most labs are designed to analyze only one (or possibly two) of these classes. Another barrier to limited lab resources is the ability to integrate data from multiple similar investigations to test the veracity of hypotheses that may be generated from omic studies. In their review, Mantini et al. describe various methods for integrating phosphoproteomic data particularly in the context of cancer research.[9] They proceed to present various methods that enable phosphoproteomic data to be integrated with data from other omic fields to provide a broader systems biology perspective. As briefly mentioned in the initial article of this special edition, a number of researchers are working towards generating omics data on single cell as a means of avoiding the inherent heterogeneity of cell populations. In an application of this area, Zhiyuan et al. developed a novel integrated microfluidic chip capable of isolating and characterizing the RNA expression profile of a single circulating cancer cells (CTCs).[10] While CTCs are effective for monitoring tumor metastasis, they are also heterogeneous. To validate their device, Zhiyuan et al. performed transcriptome profiling of single MCF-7 and white blood cells. When the transcriptomic profiles of the two cell types were compared significant genetic differences were observed based on three different measures. While proteomic diversity is mostly a result of cell-mediated post-translational modifications and the post-transcriptional occurrence of splice variants, there are many exogenous compounds that can modify proteins within the proteome.[11] In an article illustrating how dynamic the proteome can be, Tocmo et al. review how plant-derived natural products can covalently modify proteins.[13] In their article they describe how various proteomic techniques are used to identify various modifications derived from plant-based natural products can alter protein conformation, resulting in an alteration of their function and the biological activity that they modulate. One of the major benefits of the omics-era has been the discovery of the vast diversity of the biomolecules that make up cells. As scientists develop more creative uses of technologies, the number of known variations of biomolecules continues to increase. In their article, Muroski et al. build upon the observation that cyclized immonium ions that are observed in the tandem mass spectrometry (MS) spectra of acylated lysine residues could be used to identify acyl-lysine modifications.[14] They developed a stepped collision energy method that provides unique features in the tandem MS spectra of acyl-lysine containing peptides by optimizing the abundance of the immonium ion without sacrificing the information available to identify the modified peptide's sequence. Testing this novel MS method on a proteome extracted from Synotrophus aciditrophicus that had been spiked with three peptides that differed in their type and location of acetylation. This novel stepped collision energy method was able to successfully identify each of the three peptides within a complex background of peptides. In a further application, the researchers were able to identify, for the first time, a hydroxypimelyl-lysine residues within the Synotrophus aciditrophicus proteome. There is a basic paradigm in protein science that structure dictates function. But what about proteins that do not have a unique stable structure? So-called intrinsically disordered proteins (IDPs) are known to affect many cellular processes and have even been associated with diseases. The challenge, however, has been identifying and deciphering the mechanisms of this atypical class of proteins. Conventional techniques such as X-ray crystallography and nuclear magnetic resonance spectroscopy are designed to characterize stable structures within proteins but generally fail when dealing with unstructured domains. The article not only highlights the impact that MS has had on characterizing IDPs but discusses the future potential of this technique in uncovering further facets of this underexplored group of proteins.[15] Oxylipins are a family of oxygenated natural products derived from polyunsaturated fatty acids by the action of cyclooxygenase (COX) and lipoxygenase enzymes.[15] These highly potent molecules stimulate a variety of biological responses including inflammation, vasoconstriction, vasodilation, blood-clotting, etc. in response to tissue injury. In a combined targeted proteomics/metabolomics approach Scheb et al. developed a selected reaction monitoring-MS assay to measure the absolute abundance of the COX-2 enzyme and compared changes in its level to changes in the abundance of a panel of oxylipins.[16] They applied this method to study the association between COX-2 and oxylipin levels in the human colon carcinoma cell lines and lipopolysaccharide-stimulated macrophages. This combined, targeted omics approach has the potential to be expanded to include additional enzymes involved in oxylipin synthesis and can be used to further elucidate the biological mechanisms that modulate the response to tissue injury from exogenous and endogenous sources. I would like to express my deepest gratitude to all of the contributors who made this special issue possible. This past year is unlike any other that we have lived through. Even with all of the changes each of the contributors had to make to their labs and day-to-day duties because of the COVID-19 pandemic, they were generous enough to share their time to prepare an article for this issue. My hope as editor is that the reader will be impressed by the progress made in omics research in the recent past and be excited about what the next few years of continuing developments will reveal about the human organism. The author declares no conflict of interest.

• Social ranking for feature selection in machine learning models • Comparison with methods based on the Shapley value • Axiomatic and experimental analysis of solutions for feature selection • Approximation of a lexicographic social ranking solution Various methods based on the Shapley value have enjoyed notable success in recent years within the field of Explainable AI (XAI), in particular as feature selection mechanisms and for providing feature attributions for explaining machine learning models. Nevertheless, recent studies have raised concerns regarding the use of the Shapley value in this framework. In this paper, we delve deeper into these limitations through the lens of the axiomatic analysis of the Shapley value and its implications in the realm of machine learning. Leveraging on specific examples of classification models, we compare the effects of axioms for the Shapley value with other axioms for ranking methods based on a coalitional framework, where features are the “players” and the worth of a coalition of features corresponds to their predictive capacity. As an alternative feature selection method we pay particular attention to the lex-cel, a social ranking solution introduced in the recent literature at the intersection between coalitional games and social choice theory. Our analysis suggests that axioms characterizing the lex-cel, under certain circumstances, are more suitable for ranking features in machine learning models, compared to axioms satisfied by the Shapley value. Furthermore, through experiments conducted on public datasets, we show that the lex-cel outperforms some commonly employed feature selection algorithms based on the Shapley value, in particular with respect to the capacity of selecting less redundant features. An approximated version of the lex-cel, showing a satisfactory compromise between scalability of the approach and selection performance, is also presented and discussed.

The first industrial revolution focused on machines, the second one was data-centric - a third revolution combining the power of devices and information has just started and transforms our understanding of life itself. Thereby novel sensors and networks from wearable biometric devices to lab-on-a-chip platforms for exploratory fundamental research on single-biomolecule characterization and design occupy a key role. In combination with recent advances in big data analytics for life sciences, healthcare and genomics such sensors are essential tools for moving from fast and cheap personalized DNA-sequencing via smart genomics towards one-off prevention and treatment plans. Replacing state-of-the-art, one-fits-all approaches, this paradigm shifting individual assess & response scheme commonly referred to as precision medicine merges biomedical engineering, systems biology, systems genomics, and information technology. Integrated sensors for isolating, investigating and eventually manipulating single biomolecules are important experimental tools for developing next-generation DNA-sequencing platforms and for conducting ‘omics research which is a defining part of systems biology. In that context resistive pulse sensing has emerged as a powerful technology at the intersection of biotechnology and nanotechnology allowing electrical, label-free screening of biological compounds such as proteins or DNA with single-molecule, single-nucleotide and even single binding site resolution. Resistive pulse sensing technology has been at the center of recent commercial $100Ms investments in the next-generation DNA-sequencing sector. While next-generation sequencing platforms based on resistive pulse sensing techniques will mature further, the technology is also increasingly used for screening other biomolecules such as for example proteins. This allows for developing novel diagnostics and ultra-high throughput pre-clinical drug screening systems which might help to transform the pharma pipeline similarly to how the $1000-genome has revolutionized DNA-sequencing.

Scheduling divisible loads in distributed systems is the subject of divisible load theory (DLT). In this paper we show that coalitional game theory is a natural fit for modeling DLT as the participants in the scheduling algorithm must cooperate in order to execute a job. We devise a coalitional scheduling game in which the job owners and the independent organizations that own processors form coalitions in order to maximize their profits. We examine the payoffs to the participants and show that the core of the proposed coalitional scheduling game is non-empty. Then we examine the fair sharing of the payoffs among the participants using the Shapley value. Finally we study by simulation the properties of the proposed coalitional scheduling game considering different distributed systems configurations.

A graph theory and coalitional game theory-based pre-positioning of movable energy resources for enhanced distribution system resilience

In this paper, we focus on limitations in the use of the Shapley value within the field of eXplainable AI (XAI) through the lens of the axiomatic analysis and its implications in the realm of machine learning. As an alternative to the Shapley value, we analyse the properties of the lex-cel, a social ranking solution introduced inthe recent literature at the intersection between coalitional games and social choice theory, showing that axioms characterizing thelex-cel, under certain circumstances, are more suitable for ranking features in machine learning models, compared to those satisfied bythe Shapley value. Via experiments conducted on public datasets, we also show that the lex-cel outperforms a commonly employed feature selection algorithm based on the Shapley value, in particular with respect to the capacity of selecting less redundant features

Fine-Grained Parallel and Distributed Spatial Stochastic Simulation of Biological Reactions

Purpose University, industry and government relationships, known under the Triple Helix, have been studied under various aspects. The West African region and countries have been analysed with mutual information and transmission power, two information theory-based indicators. The purpose of this paper is to portray the landscape of West African Triple Helix innovation systems using three main game theory indicators (core, Shapley value and nucleolus) with the objective to measure the synergy within the selected innovation systems. Design/methodology/approach The collaboration between university, industry and government is modelled as a three-person coalitional game. Bibliographical data of selected countries were collected from Web of Science and organised according to collaboration patterns between the three actors. The characteristic functions of the games were computed, the cores plotted, the Shapley values and the nucleoli computed. Findings Either university or government has more power to create and lead to synergy; government shows solidarity towards university and industry in most of countries; and they are joined in their efforts by industry in two countries. The core exists in all the countries meaning that all the selected innovation systems present synergy; however, the extent is limited and varies over countries. Research limitations/implications Innovation includes all research products; however, this study focuses on publications only. Originality/value Synergy within a Triple Helix innovation system is studied up to now with information theory indicators. The paper portrays the landscape of West African Triple Helix innovation systems using three main game theory indicators: the core, the Shapley value and the nucleolus and gives a new way to study university, industry and government relationships.

Book Review| December 01 1963 Review: Life and Disease, New Perspectives in Biology and Medicine, by Dwight J. Ingle Life and Disease, New Perspectives in Biology and MedicineDwight J. Ingle The American Biology Teacher (1963) 25 (8): 639. https://doi.org/10.2307/4440496 Views Icon Views Article contents Figures & tables Video Audio Supplementary Data Peer Review Share Icon Share Twitter LinkedIn Tools Icon Tools Get Permissions Cite Icon Cite Search Site Citation Review: Life and Disease, New Perspectives in Biology and Medicine, by Dwight J. Ingle. The American Biology Teacher 1 December 1963; 25 (8): 639. doi: https://doi.org/10.2307/4440496 Download citation file: Ris (Zotero) Reference Manager EasyBib Bookends Mendeley Papers EndNote RefWorks BibTex toolbar search Search Dropdown Menu toolbar search search input Search input auto suggest filter your search All ContentThe American Biology Teacher Search This content is only available via PDF. Article PDF first page preview Close Modal You do not currently have access to this content.

This chapter reviews the definition and perspectives of precision medicine versus personalized medicine relevant to research and clinical practice. It examines the existing and future technologies that could speed the development of herbal products for personalized cancer chemotherapy leading to enhanced efficacy and reduced cost for treatment of resistant cancer in individual patients. As genetic variations/mutations are often important factors relating to a particular disease as well as drug response, the first essential step for precision medicine would be the identification of such mutations in specific genes at multiple cell regulatory levels. Precision medicine can involve a single drug or combination of drugs. A good example of single drug is pembrolizumab for metastatic non-small cell lung cancer whose tumor cells express programmed death ligand 1 and progress after platinum-based chemotherapy. While precision medicine and personalized medicine are often synonymous in cancer chemotherapy, the initiating steps could be different in drug development versus practice.

In the field of life sciences, although system biology and "precision medicine" introduce some complex scientifific methods and techniques, it is still based on the "analysis-reconstruction" of reductionist theory as a whole. Adaptability of complex system increase system behaviour uncertainty as well as the difficulties of precise identifification and control. It also put systems biology research into trouble. To grasp the behaviour and characteristics of organism fundamentally, systems biology has to abandon the "analysis-reconstruction" concept. In accordance with the guidelines of complexity science, systems biology should build organism model from holistic level, just like the Chinese medicine did in dealing with human body and disease. When we study the living body from the holistic level, we will fifind the adaptability of complex system is not the obstacle that increases the diffificulty of problem solving. It is the "exceptional", "right-hand man" that helping us to deal with the complexity of life more effectively.

Coalitional game deals with the situation that involves cooperations among the players. When the payoffs are characterized by uncertain variables, classical coalitional game evolves to uncertain coalitional game. Some solutions of uncertain coalitional game have been proposed such as core and Shapley value. This paper goes further to present another concept of solution—stable set for uncertain coalitional game, and shows that the core is the subset of the stable set in an uncertain coalitional game. Finally, an electricity suppliers cooperation problem is analyzed by the stable set in uncertain coalitional game.

Synergy within a Triple Helix innovation system has been measured in relevant literature using mutual information and transmission power, all based on Shannon’s information theory. However, as a complex system, Triple Helix relationships may also be analysed with various techniques and tools from other disciplines among which game theory. Thus, the synergy may be measured with indicators like the core, the Shapley value and the nucleolus. The core measures the extent of the synergy, the Shapley value indicates an actor’s strength to lead to and create synergy and the nucleolus determines an actor’s strength to maintain synergy. The Triple Helix innovation systems of eight countries among which four developed—USA, UK, Germany and France—and four emerging—Russia, India, Brazil and China—were analysed based on their scientific output using game theory. It appears that the biggest Triple Helix science producer has more power to lead to and create synergy; government shows solidarity to maintain synergy within the innovation system. The level of synergy is higher in developing countries (led by France, 1.7–2%) than in emerging ones (led by Brazil, less than or equal to 1%), operating a division of selected countries according to their level of development. The study shows that state intervention in the economy influences the position of the core on a ternary diagram.