Uniform Convergence Research Articles

Transposable Elements Are Differentially Expressed in MDS Stem Cells in a Disease-Stage-Specific Manner

Introduction Transposable elements (TEs) are widely spread across the human genome. Although TEs have been implicated in the pathogenesis of hematological malignancies, their expression in myelodysplastic syndromes (MDS) has not yet received extensive analysis. In the present study, we performed RNA sequence analysis of bone marrow stem and progenitor cells from patients with MDS and secondary acute myeloid leukemia (AML) and profiled differentially expressed TEs. Methods Bone marrow samples were obtained from each patient at the time of diagnosis. Normal bone marrow (NBM) samples from four healthy individuals were purchased from Lonza and processed in the same manner. After separation of CD34 + cells using magnetic beads, we purified lineage marker-negative (Lin -) CD34 + CD38 - stem cells and Lin - CD34 + CD38 + progenitor cells by cell sorting. The sorted cells were subjected to RNA-seq analysis using oligo-dT primers, which can capture poly-adenylated TE transcripts. Each read was aligned to the reference genome, then the annotation of the transcripts was performed using RepeatMasker, a program that screens DNA sequences for interspersed repeats and low-complexity DNA sequences (https://www.repeatmasker.org). Read counts for individual TEs were aggregated to the level of TE class, family, and subfamily according to hierarchies defined by the human RepeatMasker (repClass, repFamily, and repName, respectively). To generate raw counts of individual TE regions in RepeatMasker, the bedtools map function (version 2.27.1) was used with bed files of each sample. Normalization and significant expression differences were detected using DESeq2 (version 2.2.1). Differentially expressed TEs were defined using the cutoff of adjusted P values (q < 0.05). DESeq2 normalized counts were obtained from read counts by considering sequence depth, gene length, and RNA composition. Normalized counts were z-score scaled, and then subjected to uniform manifold approximation and projection (UMAP). Results Thirty-two patients were included in the present study. According to the 2016 revision to the World Health Organization classification (Arber DA, et al. Blood. 2016; 127: 2391-2405), 13 patients (40.6%) were diagnosed with MDS with multilineage dysplasia (MDS-MLD), 15 patients (46.9%) with MDS with excess blasts (MDS-EB), and 4 patients (12.5%) with AML with myelodysplasia-related changes (AML-MRC). We found that 26 and 17 repName TEs were significantly up-regulated in stem and progenitor cells compared with their NBM counterparts, and the majority of them belonged to LTRs, followed by DNA transposons. Since each LTR has multiple individual regions where similar sequences are scattered, LTR expression was further distinguished by the genomic locations of individual LTRs. Comparison of MDS-MLD, MDS-EB, and AML/MRC stem and progenitor cells to NBM counterparts revealed that more LTRs were differentially expressed in stem cells than in progenitor cells. Among them, many of the up-regulated LTRs were shared among different disease types (MDS-MLD, MDS-EB, and AML-MRC) in both stem and progenitor cells. AML-MRC had the largest number of differentially up- and down-regulated LTRs in stem cells. MDS-EB had the largest number of differentially up-regulated LTRs in progenitor cells, while only a small portion of LTRs were down-regulated in progenitor cells. LTRs at individual regions were differentially expressed in a more disease-specific manner, particularly in stem cells. Accordingly, UMAP plots that visualized the expression profiles of LTRs at individual regions showed that AML-MRC stem cells form a unique cluster, discriminating AML-MRC from MDS. To further understand the transcriptional reprogramming of LTRs during the disease progression from MDS to secondary AML, we analyzed the expression of differentially expressed LTRs at individual regions specific to AML-MRC by K-means clustering. They were subdivided into four clusters, of which Clusters 1, 3, and 4 showed gradual upregulation during disease progression, while Cluster 2 showed comparable upregulation at MDS and AML-MRC stages. Conclusions Our results suggest that specific LTRs behave in a disease-status and cell-type-specific manner in MDS and during leukemic transformation, providing fundamental information for understanding the pathogenesis of MDS.

On convergence and singularity of conditional copulas of multivariate Archimedean copulas, and conditional dependence

The present paper derives an explicit expression for (a version of) every uni- and multivariate conditional distribution (i.e., Markov kernel) of Archimedean copulas and uses this representation to generalize a recently established result, saying that in the class of multivariate Archimedean copulas standard uniform convergence implies weak convergence of almost all univariate Markov kernels, to arbitrary multivariate Markov kernels. Moreover, it is proved that an Archimedean copula is singular if, and only if, almost all uni- and multivariate Markov kernels are singular. These results are then applied to conditional Archimedean copulas which are reintroduced largely from a Markov kernel perspective and it is shown that convergence, singularity and conditional increasingness carry over from Archimedean copulas to their conditional copulas. As a consequence, the surprising fact is established that estimating (the generator of) an Archimedean copula directly yields an estimator of (the generator of) its conditional copula. Building upon that, we sketch the use and estimation of a conditional version of a recently introduced dependence measure as alternative to well-known conditional versions of association measures in order to study the dependence behavior of Archimedean models when fixing covariate values.

ReaLigands: A Ligand Library Cultivated from Experiment and Intended for Molecular Computational Catalyst Design.

Computational catalyst design requires identification of a metal and ligand that together result in the desired reaction reactivity and/or selectivity. A major impediment to translating computational designs to experiments is evaluating ligands that are likely to be synthesized. Here, we provide a solution to this impediment with our ReaLigands library that contains >30,000 monodentate, bidentate (didentate), tridentate, and larger ligands cultivated by dismantling experimentally reported crystal structures. Individual ligands from mononuclear crystal structures were identified using a modified depth-first search algorithm and charge was assigned using a machine learning model based on quantum-chemical calculated features. In the library, ligands are sorted based on direct ligand-to-metal atomic connections and on denticity. Representative principal component analysis (PCA) and uniform manifold approximation and projection (UMAP) analyses were used to analyze several tridentate ligand categories, which revealed both the diversity of ligands and connections between ligand categories. We also demonstrated the utility of this library by implementing it with our building and optimization tools, which resulted in the very rapid generation of barriers for 750 bidentate ligands for Rh-hydride ethylene migratory insertion.

Differential variable speed limit control strategy consider lane assignment at the freeway lane drop bottleneck

Reduction in the number of lanes (lane drop) is typical on freeways due to road design, incidents, or road maintenance. As traffic demand increases, the bottleneck becomes activated because of there are many of vehicle lane-changing behaviors at the bottleneck. At the same time, the traffic density increases, resulting in a capacity drop phenomenon at the bottleneck. The degree of the capacity drop is also closely related to the traffic density at the bottleneck. Variable Speed Limit (VSL) is a widely used Active Traffic Management (ATM) method. In traditional research, most VSL control strategies consider all lanes uniformly and set uniform dynamic speed limits. However, traffic varies by lane. This study proposed a differential variable speed limit (DVSL) control strategy for each lane considering lane assignment. The DVSL control strategy sets the dynamic speed limit for each lane to attract a certain number of vehicles to change lanes before the bottleneck and reduce the impact of traffic capacity drop. This strategy is simulated based on extending the cell transmission model (CTM) to multi-lane. The multi-level iterative optimization algorithm obtains each lane's dynamic speed limit values. The experimental results show that the proposed DVSL control strategy can alleviate traffic congestion and improve efficiency. Total travel time (TTT) improved by 7.48 % compared to the no-control measures.

On sequences which are not uniformly converging on any open subset

We consider the property of nonuniform convergence to 0 of a sequence of functions on any open subset of a metric space. We consider three examples with respect to three different characteristics. Next we show that the three characteristics cannot be present simultaneously. For this purpose we introduce the so-called height function, which we use to quantify how far is a sequence of functions from satisfying any of the third characteristic. Moreover, we study properties of the height function and its relation to uniform convergence. Finally, we show that this quantification is precise.

Single-cell RNA sequencing of human prostate basal epithelial cells reveals zone-specific cellular populations and gene expression signatures.

Despite evidence of genetic signatures in normal tissue correlating with disease risk, prospectively identifying genetic drivers and cell types that underlie subsequent pathologies has historically been challenging. The human prostate is an ideal model to investigate this phenomenon because it is anatomically segregated into three glandular zones (central, peripheral, and transition) that develop differential pathologies: prostate cancer in the peripheral zone (PZ) and benign prostatic hyperplasia (BPH) in the transition zone (TZ), with the central zone (CZ) rarely developing disease. More specifically, prostatic basal cells have been implicated in differentiation and proliferation during prostate development and regeneration; however, the contribution of zonal variation and the critical role of basal cells in prostatic disease etiology are not well understood. Using single-cell RNA sequencing of primary prostate epithelial cultures, we elucidated organ-specific, zone-specific, and cluster-specific gene expression differences in basal cells isolated from human prostate and seminal vesicle (SV). Aggregated analysis identified ten distinct basal clusters by Uniform Manifold Approximation and Projection. Organ specificity compared gene expression in SV with the prostate. As expected, SV cells were distinct from prostate cells by clustering, gene expression, and pathway analysis. For prostate zone specificity, we identified two CZ-specific clusters, while the TZ and PZ populations clustered together. Despite these similarities, differential gene expression was identified between PZ and TZ samples that correlated with gene expression profiles in prostate cancer and BPH, respectively. Zone-specific profiles and cell type-specific markers were validated using immunostaining and bioinformatic analyses of publicly available RNA-seq datasets. Understanding the baseline differences at the organ, zonal, and cellular level provides important insight into the potential drivers of prostatic disease and guides the investigation of novel preventive or curative treatments. Importantly, this study identifies multiple prostate basal cell populations and cell type-specific gene signatures within prostate basal epithelial cells that have potential critical roles in driving prostatic diseases. © 2023 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.

Uniform approximation of 2D Navier-Stokes equations with vorticity creation by stochastic interacting particle systems

We consider a stochastic interacting particle system in a bounded domain with reflecting boundary, including creation of new particles on the boundary prescribed by a given source term. We show that such particle system approximates 2D Navier–Stokes equations in vorticity form and impermeable boundary, the creation of particles modeling vorticity creation at the boundary. Kernel smoothing, more specifically smoothing by means of the Neumann heat semigroup on the space domain, allows to establish uniform convergence of regularized empirical measures to (weak solutions of) Navier–Stokes equations.

Stability Study of Event Dynamic Systems Based on Discrete Mathematical Random Variable Modeling

This project intends to consider the stability of a nonlinear control system with unknown dynamics. The system is controlled systematically by using the reverse step method based on Lyapunov's stability principle. This paper also gives the conclusion that when the control law is in a steady state, the output signal in this state will reach uniform boundedness or asymptotic convergence. Finally, a numerical simulation is carried out through an example, and the results show that the algorithm is effective.

Approximation Properties of Extended Beta-Type Szász–Mirakjan Operators

In this article, we introduce generalized beta extension of Szacute{a}sz-integral type operators and study their approximation properties. First, we calculate the some estimates for these operators. Further, we study the uniform convergence and order of approximation in terms of Korovkin-type theorem and modulus of continuity for the space of univariate continuous functions and bivariate continuous functions in their sections.. Moreover, numerical estimates and graphical representations for convergence of one- and two-dimensional sequences of operators are studied. In continuation, local and global approximation properties are studied in terms of the first- and second-order modulus of smoothness, Peetre’s K-functional and weight functions in various functional spaces.

Exploration of Chemical Diversity in Intercellular Quorum Sensing Signalling Systems in Prokaryotes.

Quorum sensing (QS) serves as a vital means of intercellular signalling in a variety of prokaryotes, which enables single cells to act in multicellular configurations. The potential to control community-wide responses has also sparked numerous recent biotechnological innovations. However, our capacity to utilize intercellular communication is hindered due to a scarcity of complementary signalling systems and a restricted comprehension of interconnections between these systems caused by variations in their dynamic range. In this study, we utilize uniform manifold approximation and projection and extended-connectivity fingerprints to explore the available chemical space of QS signalling molecules. We investigate and experimentally characterize a set of closely related QS signalling ligands, consisting of N-acyl homoserine lactones and the aryl homoserine lactone p-coumaroyl, as well as a set of more widely diverging QS ligands, consisting of photopyrones, dialkylresorcinols, 3,5-dimethylpyrazin-2-ol and autoinducer-2, and define their performance. We report on a set of six signal- and promoter-orthogonal intercellular QS signalling systems, significantly expanding the toolkit for engineering community-wide behaviour. Furthermore, we demonstrate that ligand diversity can serve as a statistically significant tool to predict much more complicated ligand-receptor interactions. This approach highlights the potential of dimensionality reduction to explore chemical diversity in microbial dynamics.

Exploration of Chemical Diversity in Intercellular Quorum Sensing Signalling Systems in Prokaryotes

AbstractQuorum sensing (QS) serves as a vital means of intercellular signalling in a variety of prokaryotes, which enables single cells to act in multicellular configurations. The potential to control community‐wide responses has also sparked numerous recent biotechnological innovations. However, our capacity to utilize intercellular communication is hindered due to a scarcity of complementary signalling systems and a restricted comprehension of interconnections between these systems caused by variations in their dynamic range. In this study, we utilize uniform manifold approximation and projection and extended‐connectivity fingerprints to explore the available chemical space of QS signalling molecules. We investigate and experimentally characterize a set of closely related QS signalling ligands, consisting of N‐acyl homoserine lactones and the aryl homoserine lactone p‐coumaroyl, as well as a set of more widely diverging QS ligands, consisting of photopyrones, dialkylresorcinols, 3,5‐dimethylpyrazin‐2‐ol and autoinducer‐2, and define their performance. We report on a set of six signal‐ and promoter‐orthogonal intercellular QS signalling systems, significantly expanding the toolkit for engineering community‐wide behaviour. Furthermore, we demonstrate that ligand diversity can serve as a statistically significant tool to predict much more complicated ligand‐receptor interactions. This approach highlights the potential of dimensionality reduction to explore chemical diversity in microbial dynamics.

EPCO-48. THE SINGLE-CELL AND SPATIAL TRANSCRIPTOMIC ARCHITECTURE OF MENINGEAL SOLITARY FIBROUS TUMORS PHENOCOPIES CEREBRAL VASCULAR DEVELOPMENT AND HOMEOSTASIS

Abstract INTRODUCTION Meningeal solitary fibrous tumors (SFTs) are rare mesenchymal neoplasms that are associated with local recurrence and distant metastasis. The cell states and spatial transcriptomic architecture of SFTs are unknown. METHODS Single-cell RNA sequencing or spatial transcriptomic sequencing were performed on 40,022 cells or 23,682 spatial transcriptomes from 12 meningeal SFT samples. Datasets were corrected for batch effects using Harmony and integrated using uniform maniform approximation and projection (UMAP). Clusters were defined using automated cell type classification, cell signature gene sets, cell cycle analysis, and differentially expressed marker genes. Trajectory analyses were performed using RNA velocity and pseudotime. Cell-cell communication analysis was performed using CellChat. RESULTS were validated using immunofluorescence and immunohistochemistry and generalized using comparisons to 30,934 single-cell transcriptomes from 6 meningioma samples, 139,134 perinatal human brain vascular single-cell transcriptomes, 84,138 adult human brain vascular single-cell transcriptomes, and DNA methylation profiles from 8 SFT and 221 meningioma samples. RESULTS Unsupervised hierarchical clustering of DNA methylation profiles revealed molecular distinction of meningeal SFTs from meningiomas. Deconvolution of vascular single-cell types showed SFT cell clusters resembling perinatal or adult fibroblasts, perinatal mitotic endothelia, and adult venous or arterial endothelia. UMAP and trajectory analyses showed SFTs were comprised of 8 interchangeable tumor cell states that were associated with cell adhesion (VCAM1, NCAM2), cell stress (EGR1), cell signaling (NOTCH3), extracellular matrix remodeling (ECM) (ADAMTS6, PLCG2), or protein synthesis pathways (RPL27A, RPL37). Cell-cell communication analyses identified single-cell and spatial interactions between VCAM1- or NCAM2-expressing SFT cells and endothelia or immature neurons in the tumor microenvironment, respectively. Single-cell deconvolution demonstrated evolution of ECM and protein synthesis SFT cells in paired primary/recurrent samples, and evolution of ECM SFT cells and macrophages in paired primary/metastatic samples. CONCLUSIONS The single-cell and spatial transcriptomic architecture of meningeal SFT reveals dynamic cell states that phenocopy cerebral vascular development and homeostasis.

Demystifying dimensionality reduction techniques in the 'omics' era: A practical approach for biological science students.

Dimensionality reduction techniques are essential in analyzing large 'omics' datasets in biochemistry and molecular biology. Principal component analysis, t-distributed stochastic neighbor embedding, and uniform manifold approximation and projection are commonly used for data visualization. However, these methods can be challenging for students without a strong mathematical background. In this study, intuitive examples were created using COVID-19 data to help students understand the core concepts behind these techniques. In a 4-h practical session, we used these examples to demonstrate dimensionality reduction techniques to 15 postgraduate students from biomedical backgrounds. Using Python and Jupyter notebooks, our goal was to demystify these methods, typically treated as "black boxes", and empower students to generate and interpret their own results. To assess the impact of our approach, we conducted an anonymous survey. The majority of the students agreed that using computers enriched their learning experience (67%) and that Jupyter notebooks were a valuable part of the class (66%). Additionally, 60% of the students reported increased interest in Python, and 40% gained both interest and a better understanding of dimensionality reduction methods. Despite the short duration of the course, 40% of the students reported acquiring research skills necessary in the field. While further analysis of the learning impacts of this approach is needed, we believe that sharing the examples we generated can provide valuable resources for others to use in interactive teaching environments. These examples highlight advantages and limitations of the major dimensionality reduction methods used in modern bioinformatics analysis in an easy-to-understand way.

Unsupervised machine learning classification of Fermi gamma-ray bursts using spectral parameters

ABSTRACT A thorough analysis of 2297 gamma-ray bursts (GRBs) in the Fermi catalogue is performed by using unsupervised machine learning algorithms in this paper. In our analysis, for two spectral parameter samples, namely for the peak-flux and time-integrated spectral fits, two dimensionality reduction algorithms, t-distributed stochastic neighbour embedding (t-SNE), and uniform manifold approximation and projection (UMAP), are used to generate four embedding maps; further, K-means algorithm is used for searching for the optimal clustering on the four maps. Our results show that Fermi GRBs can be well separated into two groups. For the time-integrated spectral parameters, both UMAP and t-SNE algorithms classify 372 bursts as short GRBs and 1925 bursts as long GRBs, and 384 bursts as short GRBs and 1913 bursts as long GRBs for the peak-time spectral parameters. This new classification method differs from traditional long and short classifications because it is not based on duration. In addition, it is found that the classification results of 11 GRBs are inconsistent between the integrated and peak-time spectral samples. GRB200826A is the first confirmed short GRB of collapsar origin, and the physical origins of these GRBs may be similar to it.

The impact of liver fibrosis on the progression of hepatocellular carcinoma via a hypoxia-immune-integrated prognostic model

The impact of liver fibrosis on the deterioration of hepatocellular carcinoma (HCC) remains controversial. We hope to explore this issue through establishing a fibrosis-hypoxia-glycolysis-immune related prognostic model. Liver fibrosis-related genes from Molecular Signatures Database were used to evaluate the degree of fibrosis in HCC patients from the TCGA database. The patients were divided into two groups using the fibrosis-related expression matrix based on the algorithm uniform manifold approximation and projection (UMAP) and evaluated for fibrosis by UMAP cluster and gene enrichment analysis. Prognostic model was constructed by differential analysis, LASSO, and multivariate regression analysis. Immune-infiltration analysis was performed by CIBERSORT. Quantitative PCR and immunohistochemistry were performed to measure the gene expression levels in HCC patients from our hospital. In 365 HCC patients from the TCGA database, 111 HCC patients with high fibrosis score have a worse prognosis than those with low fibrosis based on 129 genes related to liver fibrosis, which may be caused by the interaction between fibrosis, angiogenesis, hypoxia, glycolysis, inflammatory response, and high immune infiltration. We constructed a Fibrosis-Hypoxia-Glycolysis-Immune Prognostic Model (FHGISig), which could significantly predict disease progression in HCC patients. Furthermore, we revealed a close correlation between FHGISig and immune cell infiltration level as well as immune checkpoints. Finally, PCR results found TFF3 mRNA was significantly lower in cirrhotic HCC patients compared with non-cirrhotic ones. Liver fibrosis is a poor-prognostic factor for HCC, and our FHGISig could significantly predict disease progression, which could also be a potential predictive marker for immunotherapy in HCC patients.

A novel parallel multi-harmonic global multi-channel control algorithm for helicopter active vibration control

The vibrations of the helicopter fuselage are mainly induced by the rotor with blade pass frequency and the high order harmonics. The conventional filtered-x least mean square (FxLMS) control algorithm usually used a uniform convergence factor for all the harmonics which leads poor effect in controlling multi-harmonic vibration. In this paper, a novel parallel multi-harmonic global multi-channel (MHGMC) control method is proposed to suppress the multi-harmonic vibration components transmitted from the main rotor to the fuselage. Bandpass filters are used to separate multiple frequency components of the error signals, the control channel models are identified for the corresponding frequency range to independently control each frequency component, and the reconstruction of the reference input signals can be obtained from the separated error signal, effectively improving the performance of tracking interference signals. In addition, the coupling of the control channels is considered, and the global vibration optimization is taken as the control objective to update the weight coefficient of the controller. An active vibration control system of the helicopter main gearbox eight-strut installation configuration has been established. The simulation and experiment results show that the proposed controller has faster control convergence speed and better control effect of multiple harmonics than the conventional filtered-x least mean square algorithm, and the attenuations of the disturbing frequency harmonic components at all measurement points in the experiment are over 38dB. The experimental results of variable excitation also verify that the proposed control algorithm has excellent adaptability and strong robustness.

Machine Learning-Facilitated Cross-Panel Automatic Classification at Both Sample and Cell Level in Acute Myeloid Leukemia

Introduction Flow cytometry (FC) immunophenotyping is essential for accurate and prompt diagnosis of acute myeloid leukemia (AML), with up to 95% of AML patients showing detectable aberrant immunophenotype (IP) by FC. Current manual and subjective analysis of clinical FC data, however, lacks reproducibility and leads to interpretation variation. In previous studies, we showed that supervised machine learning (ML) approaches can detect IP abnormalities and identify the acute leukemia subclassification accurately at the specimen level. We also found that comparable classification performance can be achieved using only a subset of markers of the reagent panel (Ko et al., 2018; Monaghan et al., 2022). To interpret the ML classification results, we have developed data clustering and discriminative learning methods to identify and visualize the diagnostic cell populations in a 2D space for hematopathology review (Ji et al., 2020; Lee et al., 2018). Built upon these approaches, this study reports our findings on developing an interpretable cross-panel ML classification model to support timely and accurate diagnosis and subclassification of AML. Materials and Method The AML diagnosis model at the sample level utilized FC data obtained from 53 bone marrow (BM) aspirate samples from Roswell Park Comprehensive Cancer Center (RPCCC) and 50 BM aspirate samples from UPMC. Data from RPCCC were acquired using a Beckman Coulter Navios EX and measured with the ClearLLab10C panel; while data from UPMC were acquired using a BD FACSCantoII and measured with a UPMC-developed diagnostic panel. We used only the common parameters collected within these 103 FC data sets and applied a previously published Gaussian Mixture Model-Support Vector Machine (GMM-SVM) based approach to train a cross panel sample-level classification model for AML diagnosis (Ko et al., 2018; Monaghan et al., 2022). Three-fold cross-validation was conducted to assess the model performance, using sensitivity, specificity, classification accuracy, and area under the receiver operating characteristic curve (AUC). For the cell-level visualization and analysis, all samples in the training set were pooled together, before a uniform manifold approximation and projection (UMAP) process was applied to transform and visualize the cell populations within the samples. Specifically, 26 non-neoplastic and 27 AML samples from the M2 panel of the ClearLLab10C from the RPCCC cohort were combined for the UMAP visualization. Then the FlowSOM unsupervised clustering method and the DAFI semi-supervised clustering method (Lee et al., 2018) were applied to identify all of the cell populations that have significant differences between the AML samples and the non-neoplastic controls. Results The cross-panel AML versus non-neoplastic classifier, combining FC data from UPMC and RPCCC achieved exceptional performance. The AML detection model using all parameters on the ClearLLab10C panel achieved an AUC of 100%. When using only the overlapping parameters between the ClearLLab10C and the UPMC panels, the classification model achieved an AUC of 99.5%, while the cross-panel classifier achieved an AUC of 99.1% (Table 1). Specimen-level visualization confirmed the separation of AML versus non-neoplastic controls. The UMAP plots, created by pooling non-neoplastic and AML samples, exhibited distinct patterns for different AML subtypes (M1 to M5) and a genetic mutation (e.g., AML IDH1), which were distinct from control samples. Using the automated gating analysis, cell populations identified as different between AML and control samples were characterized and visualized in 2D-by-2D scatter plots for straightforward interpretation and assessment based on their immunophenotype. Conclusion Our study demonstrated the feasibility to develop and assess a ML classification model for AML detection and diagnosis using overlapping markers in FC data across reagent panels and sites. Advanced visualization and automated gating analysis enabled the interpretation of the ML classification by identifying and characterizing the diagnostic cell populations and their phenotypic heterogeneity, providing insights into AML heterogeneity at both the specimen and the single cell level. These findings highlight the potential of ML capability for improving patient care by timely and accurate diagnosis and subclassification of AML.

Network Modeling of Apoptosis-Related Genes Reveal a High-Risk Cluster in Multiple Myeloma

The plasma cell cancer multiple myeloma (MM) varies significantly in genomic characteristics, response to therapy, and long-term prognosis. Recent studies have shown the complexity of biological systems associated with poor outcomes in MM. One hallmark of cancer pathogenesis, applicable to MM, is the dysregulation of apoptotic cell death. Understanding the relationship between known drivers of MM and critical pathways in the apoptosis gene network is key towards identifying novel therapeutic targets. Recently, there has been a surge of interest in developing network-based approaches to study gene interactions and use machine learning tools to further elucidate useful biological factors. It is well-known that genes operate as network systems; however, interpretable computational methods are still limited. In this work, we combine a set of network analysis tools to provide novel biological insights into how the apoptosis network is dysregulated in MM. We identify a set of genes associated with poor outcomes in MM using both univariate and network-based approaches. Genomic and clinical data, including RNA-sequencing (RNA-seq) and copy number alteration (CNA), for 659 subjects were obtained from the Multiple Myeloma Research Foundation's CoMMpass (IA19) database. The apoptosis gene set is defined using the apoptosis hallmark pathway provided by the gene set enrichment analysis tool. The interactions between the given list of genes are provided via MetaCore. Univariate gene modeling is done using a Cox proportional hazards model. To determine the relationship between genes, we used Ollivier-Ricci curvature (ORC), a measure of network robustness. The ORC value between two genes incorporates information about both the two genes in question and their local neighborhoods, a concept critical for understanding the role a neighborhood of genes has on a given connection. Once edge information is computed, we used a uniform manifold approximation and projection (UMAP) to reduce the complexity of the dataset. The resulting projection is clustered using k-means clustering, and survival analysis of the given clusters is done using the Kaplan-Meier (KM) method. To understand which features are most informative, we used a random forest to predict the cluster labels using the RNA-seq data. After five-fold cross-validation, feature importance was computed using random forest's permutation importance methodology. All statistical significance testing was corrected for multiple comparisons via the Benjamini-Hochberg false discovery rate method with an alpha set to 0.05. Using RNA-seq data of genes associated with apoptosis, we identified nineteen apoptosis genes associated with poor prognosis in MM. Some of these genes, such as ANXA1 and WEE1 are known to have an effect in MM, while others, such as AVPR1A have yet to be studied extensively for their role in MM. When examining CNA data associated with apoptosis, a similar pattern emerged - nineteen genes were associated with poor prognosis. However, between the two key gene sets found by RNA-seq and CNA, only FAS appeared in both lists. FAS, Fas Cell Surface Death Receptor, is critical for triggering apoptosis. Half of the 37 genes identified using RNA-seq and CNA data are directly connected to the TP53 gene. While TP53 mutations are not a marker of the presence of MM itself, TP53 mutations have been shown to be associated with poorer outcomes. After reducing the complexity of the Ollivier-Ricci curvature analysis of the RNA-seq apoptosis network using UMAPs, the data revealed five clusters with KM curves for progression-free survival were significant (p<0.001). To identify the genes most impactful for separating these clusters, we trained a random forest model on the RNA-seq - note, it was the edge robustness values that were clustered - and evaluated its performance for predicted class membership. The model had an overall accuracy of approximately 78% with F1 scores for each class ranging from 61% to 87%. The top three features associated with the model include CTNNB1, RELA, and LEF1. Of note, CTNNB1 and RELA are directly interacted with the TP53 gene. Because these three genes discriminate between all five clusters, their expression levels may play a mediating effect on MM outcomes. Further investigation with external datasets is needed to validate these results.

High Dimensional Detection of Non-Malignant B-Cells and Its Clinical Implications in Patients with Chronic Lymphocytic Leukaemia (CLL) Undergoing Frontline Therapy

Introduction: The analysis of non-malignant B-cells in chronic lymphocytic leukaemia (CLL) is challenging due to their scarcity and paucity of reliable identification methods. However, given the well-defined immunophenotypic characteristics of CLL cells, it is feasible to distinguish and quantify non-malignant B-cells by incorporating an extensive panel of B-cell markers into unsupervised high-dimensional analysis. In this study, we performed high dimensional mass cytometry analysis of circulating B-cells in patients enrolled in the NCRI phase III RIAltO trial, which compared frontline chemoimmunotherapy (CIT) regimens in CLL. Methods: Mass cytometry was performed on samples obtained from CLL patients enrolled in the RIAltO trial (NCT01678430). Patients were randomly assigned to receive bendamustine or chlorambucil plus ofatumumab, with or without idelalisib. Samples were collected before treatment and at early (median 11.2 [IQR 11 - 11.7] months) and late (18.5 [16.1 - 22.6] months) post-treatment timepoints, as well as at disease relapse. Healthy controls (HC) were included for comparison. CD19 + B cells were analysed using a customised antibody panel with 29 surface-antigen markers. Results: A total of 198 samples from 79 patients and three HC were analysed. Unsupervised analysis using uniform manifold approximation and projection (UMAP) and FlowSOM identified two distinct groups of CD19 + B-cell clusters. Malignant clusters were identified based on established immunophenotypic markers of CLL, while non-malignant clusters were identified using the inverse co-expression of these markers (CD5 -CD43 -ROR1 -CD20 +CD79b +CD81 +). In keeping with the assumption that non-malignant clusters comprised physiological B-cells, they were the majority cell type among HC, rare before treatment or at relapse, and variably detected following therapy. When relating non-malignant clusters with clinical outcomes, their re-emergence following therapy correlated with an increase in markers of haematopoietic recovery, including haemoglobin concentration (R 2 = 0.4, p = 0.002), neutrophil count (R 2 = 0.39, p = 0.002) and platelet count (R 2 = 0.41, p = 0.002). Furthermore, re-emergence of non-malignant, CD27 + memory B-cells the at late post-treatment timepoint positively correlated with serum immunoglobulins (Ig): IgA (R 2 = 0.41, p = 0.012), IgM (R 2 = 0.37, p = 0.027) and IgG (R 2 = 0.42, p = 0.01) levels. Notably, stratification of patients based on post-treatment median cluster size showed that a higher frequency of non-malignant B-cells correlated with significantly longer time to progression (TTP; median 49.9 [95% CI: 28.5 - NR] versus 9.7 [7.2 - 26.8] months, p < 0.001). This correlation remained significant after adjusting for baseline characteristics, CLL prognostic factors, treatment allocation, minimal residual disease (MRD) status, and haematopoietic recovery. Subsequently, a Boolean gating strategy which prospectively defines non-malignant B-cells (CD5 -CD43 -ROR1 -CD20 +CD79b +CD81 +) was devised. Despite a weak correlation with MRD (R 2 = -0.27, p = 0.084), the proportion of non-malignant B cells at the end of treatment was found to be an independent determinant of TTP and overall survival (OS) among patients with ≥ 0.1% residual CLL cells. Conclusion: To our knowledge, this is the first study to identify correlations between the re-emergence of non-malignant B-cells following anti-CLL therapy and specific clinical endpoints such as haematopoietic recovery, serum Ig levels, TTP and OS. Although further validation is required in the context of targeted agents, our findings suggest that therapy-induced re-emergence of non-malignancy B-cells may have important biological and clinical implications in CLL.

The Attractors of Super Iterated Function Systems on Fractal Space

Fractal geometry is a new branch of mathematics, which has a very wide and important application in everyday life. Iterated function systems (abbrev: IFS ) play an important role in constructing fractal set, which is a critical method in fractal geometry. The attractor is also an important concept of fractal geometry, which is obtained by an iterated function systems or a super iterated function systems (abbrev: SIFS). The SIFS is a significant generalization of the IFS by taking the number of functions of the IFS from finity to infinity. This study extend IFS to SIFS through uniform convergence. Then, some important properties similar to IFS are obtained, particularly the property of the attractors. Finally, The relationship among the attractors between SIFS and IFS is obtained, denoted by .