다중혈관 관상동맥 환자에서 y-문합을 이용하여 양쪽 내흉동맥만을 사용한 우회술의 조기 성적

Chronic inflammation is characterized by continuous recruitment and activation of immune cells such as monocytes in response to a persistent stimulus. Production of proinflammatory mediators by monocytes leads to tissue damage and perpetuates the inflammatory response. However, the mechanism(s) responsible for the sustained influx of monocytes in chronic inflammation are not well defined. In chronic peritonitis induced by pristane, the persistent recruitment of Ly6C(hi) inflammatory monocytes into the peritoneum was abolished in type I interferon (IFN-I) receptor-deficient mice but was unaffected by the absence of IFN-gamma, tumor necrosis factor-alpha, interleukin-6, or interleukin-1. IFN-I signaling stimulated the production of chemokines (CCL2, CCL7, and CCL12) that recruited Ly6C(hi) monocytes via interactions with the chemokine receptor CCR2. Interestingly, after 2,6,10,14-tetramethylpentadecane treatment, the rapid turnover of inflammatory monocytes in the inflamed peritoneum was associated with a lack of differentiation into Ly6C(lo) monocytes/macrophages, a more mature subset with enhanced phagocytic capacity. In contrast, Ly6C(hi) monocytes differentiated normally into Ly6C(lo) cells in IFN-I receptor-deficient mice. The effects of IFN-I were specific for monocytes as granulocyte migration was unaffected in the absence of IFN-I signaling. Taken together, our findings reveal a novel role of IFN-I in promoting the recruitment of inflammatory monocytes via the chemokine receptor CCR2. Continuous monocyte recruitment and the lack of terminal differentiation induced by IFN-I may help sustain the chronic inflammatory response.

Type-I interferon modulates monocyte recruitment and maturation in chronic inflammation

We used 27 human single cell transcriptomics (SCT) data sets to develop an artificial neural network (ANN) model for classification of Peripheral Blood Mononuclear Cells (PBMC). We demonstrated that highly accurate models for the classification of PBMC subtypes can be developed by combining multiple independent data sets to form training data sets. A significant data preparation effort was needed for building predictive models. Using a data set of ∼120,000 single cell instances we showed the accuracy of classification of PBMC call of ∼ 90%. Optimization techniques and the addition of new high-quality data sets for model training are expected to improve PBMC subtype classification accuracy.

Classification of Five Cell Types from PBMC Samples using Single Cell Transcriptomics and Artificial Neural Networks

The immune system varies in cell types, states, and locations. The complex networks, interactions, and responses of immune cells produce diverse cellular ecosystems composed of multiple cell types, accompanied by genetic diversity in antigen receptors. Within this ecosystem, innate and adaptive immune cells maintain and protect tissue function, integrity, and homeostasis upon changes in functional demands and diverse insults. Characterizing this inherent complexity requires studies at single-cell resolution. Recent advances such as massively parallel single-cell RNA sequencing and sophisticated computational methods are catalyzing a revolution in our understanding of immunology. Here we provide an overview of the state of single-cell genomics methods and an outlook on the use of single-cell techniques to decipher the adaptive and innate components of immunity.

Single-cell transcriptomics to explore the immune system in health and disease

The ZX-calculus is a universal graphical language for qubit quantum computation, meaning that every linear map between qubits can be expressed in the ZX-calculus. Furthermore, it is a complete graphical rewrite system: any equation involving linear maps that is derivable in the Hilbert space formalism for quantum theory can also be derived in the calculus by rewriting. It has widespread usage within quantum industry and academia for a variety of tasks such as quantum circuit optimisation, error-correction, and education.The ZW-calculus is an alternative universal graphical language that is also complete for qubit quantum computing. In fact, its completeness was used to prove that the ZX-calculus is universally complete. This calculus has advanced how quantum circuits are compiled into photonic hardware architectures in the industry.Recently, by combining these two calculi, a new calculus has emerged for qubit quantum computation, the ZXW-calculus. Using this calculus, graphical-differentiation, -integration, and -exponentiation were made possible, thus enabling the development of novel techniques in the domains of quantum machine learning and quantum chemistry.Here, we generalise the ZXW-calculus to arbitrary finite dimensions, that is, to qudits. Moreover, we prove that this graphical rewrite system is complete for any finite dimension. This is the first completeness result for any universal graphical language beyond qubits.

Completeness for arbitrary finite dimensions of ZXW-calculus, a unifying calculus

Single cell transcriptomics (SCT) technology reveals gene expression of individual cells. Peripheral blood mononuclear cells (PBMC) are important diagnostic targets in immunology. In this study, we obtained and standardized 27 SCT data sets, derived from healthy PBMC samples using 10x SCT. We used artificial neural networks (ANN) to assess the ability of ANN to classify main PBMC cell types. Incremental learning by the gradual addition of new data sets to ANN training improved classification. The overall prediction accuracy of the final step of incremental learning reached 93% in 4-class classification.

Classification of PBMC cell types using scRNAseq, ANN, and incremental learning


 A domain-theoretic framework is presented for validated robustness analysis of neural networks. First, global robustness of a general class of networks is analyzed. Then, using the fact that Edalat’s domain-theoretic L-derivative coincides with Clarke’s generalized gradient, the framework is extended for attack-agnostic local robustness analysis. The proposed framework is ideal for designing algorithms which are correct by construction. This claim is exemplified by developing a validated algorithm for estimation of Lipschitz constant of feedforward regressors. The completeness of the algorithm is proved over differentiable networks and also over general position 
 
 
 
${\mathrm{ReLU}}$

 
 networks. Computability results are obtained within the framework of effectively given domains. Using the proposed domain model, differentiable and non-differentiable networks can be analyzed uniformly. The validated algorithm is implemented using arbitrary-precision interval arithmetic, and the results of some experiments are presented. The software implementation is truly validated, as it handles floating-point errors as well.

A Domain-Theoretic Framework for Robustness Analysis of Neural Networks

In this paper, we develop a graphical calculus to rewrite photonic circuits involving light-matter interactions and non-linear optical effects. We introduce the infinite ZW calculus, a graphical language for linear operators on the bosonic Fock space which captures both linear and non-linear photonic circuits. This calculus is obtained by combining the QPath calculus, a diagrammatic language for linear optics, and the recently developed qudit ZXW calculus, a complete axiomatisation of linear maps between qudits. It comes with a 'lifting' theorem allowing to prove equalities between infinite operators by rewriting in the ZXW calculus. We give a method for representing bosonic and fermionic Hamiltonians in the infinite ZW calculus. This allows us to derive their exponentials by diagrammatic reasoning. Examples include phase shifts and beam splitters, as well as non-linear Kerr media and Jaynes-Cummings light-matter interaction.

Light-matter interaction in the ZXW calculus

In this report we present a new modelling framework for concepts based on quantum theory, and demonstrate how the conceptual representations can be learned automatically from data. A contribution of the work is a thorough category-theoretic formalisation of our framework. We claim that the use of category theory, and in particular the use of string diagrams to describe quantum processes, helps elucidate some of the most important features of our quantum approach to concept modelling. Our approach builds upon Gardenfors' classical framework of conceptual spaces, in which cognition is modelled geometrically through the use of convex spaces, which in turn factorise in terms of simpler spaces called domains. We show how concepts from the domains of shape, colour, size and position can be learned from images of simple shapes, where individual images are represented as quantum states and concepts as quantum effects. Concepts are learned by a hybrid classical-quantum network trained to perform concept classification, where the classical image processing is carried out by a convolutional neural network and the quantum representations are produced by a parameterised quantum circuit. We also use discarding to produce mixed effects, which can then be used to learn concepts which only apply to a subset of the domains, and show how entanglement (together with discarding) can be used to capture interesting correlations across domains. Finally, we consider the question of whether our quantum models of concepts can be considered conceptual spaces in the Gardenfors sense.

Formalising and Learning a Quantum Model of Concepts

We introduce a novel compositional description of Feynman diagrams, with well-defined categorical semantics as morphisms in a dagger-compact category. Our chosen setting is suitable for infinite-dimensional diagrammatic reasoning, generalising the ZX calculus and other algebraic gadgets familiar to the categorical quantum theory community. The Feynman diagrams we define look very similar to their traditional counterparts, but are more general: instead of depicting scattering amplitude, they embody the linear maps from which the amplitudes themselves are computed, for any given initial and final particle states. This shift in perspective reflects into a formal transition from the syntactic, graph-theoretic compositionality of traditional Feynman diagrams to a semantic, categorical-diagrammatic compositionality. Because we work in a concrete categorical setting—powered by non-standard analysis—we are able to take direct advan-tage of complex additive structure in our description. This makes it possible to derive a particularly compelling char-acterisation for the sequential composition of categorical Feynman diagrams, which automatically results in the superposition of all possible graph-theoretic combinations of the individual diagrams themselves (cf. Figure 1 above).

Razin A. Shaikh

Papers

Classification of PBMC cell types using scRNAseq, ANN, and incremental learning

A Domain-Theoretic Framework for Robustness Analysis of Neural Networks

Light-matter interaction in the ZXW calculus

Formalising and Learning a Quantum Model of Concepts

Categorical Semantics for Feynman Diagrams