Sequence Of Variables Research Articles

Adaptive Quality Diagnosis Framework for Production Lines in a Smart Manufacturing Environment

Production lines in manufacturing environments benefit from quality diagnosis methods based on learning techniques since their ability to adapt to the runtime conditions improves performance, and at the same time, difficult computational problems can be solved in real time. Predicting the divergence of a product’s physical parameters from an acceptable range of values in a manufacturing line is a process that can assist in delivering consistent and high-quality output. Costs are saved by avoiding bursts of defective products in the pipeline’s output. An innovative framework for the early detection of a product’s physical parameter divergence from a specified quality range is designed and evaluated in this study. This framework is based on learning automata to find the sequences of variables that have the highest impact on the automated sensor measurements that describe the environmental conditions in the production line. It is shown by elaborate evaluation that complexity is reduced and results close to optimal are feasible, rendering the framework suitable for deployment in practice.

Complete variable domain sequences of monoclonal antibody light chains identified from untargeted RNA sequencing data.

Monoclonal antibody light chain proteins secreted by clonal plasma cells cause tissue damage due to amyloid deposition and other mechanisms. The unique protein sequence associated with each case contributes to the diversity of clinical features observed in patients. Extensive work has characterized many light chains associated with multiple myeloma, light chain amyloidosis and other disorders, which we have collected in the publicly accessible database, AL-Base. However, light chain sequence diversity makes it difficult to determine the contribution of specific amino acid changes to pathology. Sequences of light chains associated with multiple myeloma provide a useful comparison to study mechanisms of light chain aggregation, but relatively few monoclonal sequences have been determined. Therefore, we sought to identify complete light chain sequences from existing high throughput sequencing data. We developed a computational approach using the MiXCR suite of tools to extract complete rearranged IGVL-IGJL sequences from untargeted RNA sequencing data. This method was applied to whole-transcriptome RNA sequencing data from 766 newly diagnosed patients in the Multiple Myeloma Research Foundation CoMMpass study. Monoclonal IGVL-IGJL sequences were defined as those where >50% of assigned IGK or IGL reads from each sample mapped to a unique sequence. Clonal light chain sequences were identified in 705/766 samples from the CoMMpass study. Of these, 685 sequences covered the complete IGVL-IGJL region. The identity of the assigned sequences is consistent with their associated clinical data and with partial sequences previously determined from the same cohort of samples. Sequences have been deposited in AL-Base. Our method allows routine identification of clonal antibody sequences from RNA sequencing data collected for gene expression studies. The sequences identified represent, to our knowledge, the largest collection of multiple myeloma-associated light chains reported to date. This work substantially increases the number of monoclonal light chains known to be associated with non-amyloid plasma cell disorders and will facilitate studies of light chain pathology.

Reinforcement learning with dynamic convex risk measures

AbstractWe develop an approach for solving time‐consistent risk‐sensitive stochastic optimization problems using model‐free reinforcement learning (RL). Specifically, we assume agents assess the risk of a sequence of random variables using dynamic convex risk measures. We employ a time‐consistent dynamic programming principle to determine the value of a particular policy, and develop policy gradient update rules that aid in obtaining optimal policies. We further develop an actor–critic style algorithm using neural networks to optimize over policies. Finally, we demonstrate the performance and flexibility of our approach by applying it to three optimization problems: statistical arbitrage trading strategies, financial hedging, and obstacle avoidance robot control.

Phylogenetic relationship and comparative analysis of the main Bupleuri Radix species in China.

Bupleuri Radix (Chaihu) is a famous traditional Chinese medicine derived from Bupleurum, Apiaceae. The origin of cultivated Chaihu germplasm in China is unclear, which has led to unstable Chaihu quality. In this study, we reconstructed the phylogeny of the main Chaihu germplasm species in China and identified potential molecular markers to authenticate its origin. Three Bupleurum species (eight individuals), B. bicaule, B. chinense, and B. scorzonerifolium, were selected for genome skimming. Published genomes from B. falcatum and B. marginatum var. stenophyllum were used for comparative analysis. Sequences of the complete plastid genomes were conserved with 113 identical genes ranging from 155,540 to 155,866 bp in length. Phylogenetic reconstruction based on complete plastid genomes resolved intrageneric relationships of the five Bupleurum species with high support. Conflicts between the plastid and nuclear phylogenies were observed, which were mainly ascribed to introgressive hybridization. Comparative analysis showed that noncoding regions of the plastomes had most of the variable sequences. Eight regions (atpF-atpH, petN-psbM, rps16-psbK, petA-psbJ, ndhC-trnV/UAC and ycf1) had high divergence values in Bupleurum species and could be promising DNA barcodes for Chaihu authentication. A total of seven polymorphic cpSSRs and 438 polymorphic nSSRs were detected across the five Chaihu germplasms. Three photosynthesis-related genes were under positive selection, of which accD reflected the adaptation fingerprint of B. chinense to different ecological habitats. Our study provides valuable genetic information for phylogenetic investigation, germplasm authentication, and molecular breeding of Chaihu species.

Estimating function method for nonnegative autoregressive models

A stationary sequence of nonnegative random variables generated by autoregressive (AR) models may be used to describe the inter‐arrival times between events in counting processes. Even though, several such models are available in the literature, there is no unified approach to estimate their parameters. In this paper, we propose a class of combined estimating function method to estimate the model parameters of AR models with gamma marginals. The proposed method is compared with other estimation procedures and are illustrated by simulation and data analysis.

Lacunary statistical convergence of rough triple sequence via ideals

In this paper, we have introduced the notion of the lacunary [Formula: see text]-statistical convergence of triple sequences for rough variables. In addition, we have defined lacunary [Formula: see text]-statistical Cauchy sequence of rough variables in trust space and given the lacunary [Formula: see text]-statistical completeness for trust space. Some important results have also been examined.

Abstract 250: Identifying drivers of liver cancer using CRISPR activation screening in vivo

Abstract Hepatocellular Carcinoma (HCC) is the most prevalent type of liver cancer and the third leading cause of cancer deaths worldwide. Drugs approved for first- and second-line therapy extend survival by only several months. Hence, there is still a pressing need for new and effective treatments. Sequencing technologies applied to human HCC tissues have identified hundreds of genes that undergo amplification and that may be correlated with mortality. Nevertheless, correlation does not equate to functionality. We aim to annotate gene function by employing an innovative approach with Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) and catalytically dead version of CRISPR-associated endonuclease (dCas9) in a CRISPR/dCas9 activation (CRISPRa) screen to identify driver genes of HCC tumorigenesis in mice. We deliver to the liver’s hepatocytes a gRNA library targeting genes of interest on plasmids that contain: (1) a transposon system for stable integration of DNA sequences into the cell genome, (2) Myc to drive hyperproliferation, as MYC is frequently over-expressed in human HCC, (3) synergistic activation mediator complex sequences for transcriptional activation of target genes, and (4) the gRNA sequence, the only variable sequence in the library. Using datasets from the Cancer Genome Atlas, we identified 51 genes that are both enriched and amplified in HCC patients. We included gRNAs targeting Tert, Vegfa, and Ccnd1, as control genes with known roles in driving HCC. In total, the screening library contained 290 gRNAs targeting 51 genes and 3 controls. After delivery of the plasmid library to the liver, the Myc gene provided a baseline level of oncogenic hyperproliferation, and the readout was gRNA prevalence pre- and post-proliferation via next-generation sequencing. We found that the combination of Myc and CRISPRa gRNA library led to a large number of expansive tumors in the liver as compared to Myc or the CRISPRa gRNA library only, which contained few to no tumors. gRNAs targeting control genes Ccnd1 and Vegfa were enriched as expected, validating our approach to annotate gene function during tumorigenesis. Additionally, we identified enrichment in gRNAs targeting genes Zbtb7b, Vps72, and Gba in multiple tumors, implicating a role in liver tumorigenesis. ZBTB7B and GBA expression levels are increased in patients with HCC recurrence as compared to patients in remission (data from BIOSTORM trial), while VPS72 is an unfavorable prognostic marker for HCC (data from The Human Protein Atlas). In this study, we have established a novel model of liver cancer that combines the Myc oncogene with a CRISPRa gRNA library to annotate driver genes in human HCC. Additionally, we identified several promising targets for the treatment of HCC. Ultimately, this study has the potential to impact on the care of patients with HCC in the US and worldwide. Citation Format: Alexandra Mariel Vázquez Salgado, Shirui He, Kirk J. Wangensteen. Identifying drivers of liver cancer using CRISPR activation screening in vivo [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 250.

Abstract 1878: Preclinical development of a novel bispecific mAb-Trap fusion protein, IMM2520, targeting both PD-L1 and CD47 as cancer immunotherapy

Abstract The programmed cell death protein 1 (PD-1) immune checkpoint is well known in its ability to attenuate T cell-mediated immune responses in tumors expressing its ligand, programmed death-ligand 1 (PD-L1). CD47, a cell surface molecule in the immunoglobulin superfamily, is another established target overexpressed on various malignant cells and acts as an innate inhibitory checkpoint by interrupting the phagocytosis of tumor cells and downstream activation of innate responses. It has been reported that the simultaneous blockade on both CD47 and PD-L1 checkpoints inhibited tumor growth and circulating tumor cells metastasis more potently than a single antibody inhibition or blank control in the 4T1 tumor mouse model in vivo. IMM2520 is an antibody-receptor (mAb-Trap) fusion protein consisting of recombinant human SIRPα and anti- PD-L1 antibody, thereby targeting CD47 and PD-L1 to antagonize both innate and adaptive immune checkpoint pathways. IMM2520 binds to PD-L1 on the tumor cell membrane and restores the immune response of T cells. IMM2520 also reverses the inhibitory effect of tumor cells on macrophages by blocking the binding of CD47 on the tumor cell to SIRPα on macrophages. In vitro studies have demonstrated that IMM2520 not only possesses a potent binding activity to CD47 (EC50 = 0.8 nM) and PD-L1 (EC50 = 0.09 nM) respectively, but also exhibits a strong activity in blocking PD-1/PD-L1 binding (IC50 = 0.06 - 1.0 nM). IMM2520 also showed a potent antibody-dependent cellular cytotoxicity (ADCC, EC50 = 1.54 ng/ml) against Raji-PDL1 and induced a strong antibody-dependent cellular phagocytosis (ADCP, EC50 = 0.04 nM) activity (THP-1 against Raji-PDL1). In vivo efficacy study with CT26-hPD-L1/hCD47 colon cancer model demonstrated that IMM2520 at 3 mg/kg significantly reduced tumor volume (TGITV = 90.12%, P<0.01 vs Vehicle). Interestingly, IMM2520 demonstrated a potent and significantly higher anti-tumor activities than IMM01 (a SIRPα-Fc fusion protein) and IMM2505 (also a CD47XPD-L1 mAb-Trap, but the variable region sequence of this PD-L1 antibody is identical to atezolizumab). IMM2520 at dose of 6 mg/kg completely eradicated the established tumors. In another in vivo efficacy study with MC38-hCD47/hPD-L1 colon cancer model, IMM2520 demonstrated a significant anti-tumor activity in a dose-dependent manner at doses between 2 mg/kg to 20 mg/kg. Given its potent preclinical anti-tumor activity as well as the favorable safety profile, IMM2520 may serve as a potent immunotherapy for multiple cancer types by targeting PD-L1 and CD47 on tumor cells. Citation Format: Wenzhi Tian, Song Li, Dianze Chen, Wenqi Zhu, Yanan Yang, Dandan Liu, Huiqin Guo, Chunmei Yang, Li Zhang, Wei Zhang, Xiaoping Tu, Liang Peng, Gui Zhao, Ruliang Zhang, Fan Zhang, Frank Gan. Preclinical development of a novel bispecific mAb-Trap fusion protein, IMM2520, targeting both PD-L1 and CD47 as cancer immunotherapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1878.

On Statistically Pre-Cauchy Sequences of Complex Uncertain Variables Defined by Orlicz Functions

In this treatise, we define statistically pre-Cauchy sequences of complex uncertain variable for five cases of uncertainty viz., in mean, in measure, in distribution, in almost surely and in uniformly almost surely and we confine our study to statistically pre-Cauchy sequence in mean, in measure and in distribution only. Furthermore, we establish the relationship between statistically pre-Cauchy and statistically convergent sequence by using complex uncertain variables. Finally, we initiate the study of statistically pre-Cauchy sequences of complex uncertain variables via Orlicz functions also.

Almost sure central limit theorems for the maxima of randomly chosen random variables

In this paper, we give an almost sure central limit theorem (ASCLT) version of a maximum limit theorem (MLT) with an arbitrary sequence {dn, n ≥ 1} of weighted means of max{Xk, k ∈ An}, where {Xn, n ≥ 1} is a sequence of independent random variables, and {An, n ≥ 1} is a sequence of almost surely finite random subsets of positive integers independent of {Xn, n ≥ 1}. Thus we generalize the cases considered in the literature: (i) the nonrandom version of ASCLT for the MLT; (ii) the version of ASCLT for randomly indexed MLT; and (iii) the version of maximum schema of observed and unobserved random variables. We complete the paper with illustrative examples.

Accelerated Dual Averaging Methods for Decentralized Constrained Optimization

In this article, we study decentralized convex constrained optimization problems in networks. We focus on the dual averaging-based algorithmic framework that is well-documented to be superior in handling constraints and complex communication environments simultaneously. Two new decentralized dual averaging (DDA) algorithms are proposed. In the first one, a second-order dynamic average consensus protocol is tailored for DDA-type algorithms, which equips each agent with a provably more accurate estimate of the global dual variable than conventional schemes. We rigorously prove that the proposed algorithm attains <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\mathcal {O}(1/t)$</tex-math></inline-formula> convergence for general convex and smooth problems, for which existing DDA methods were only known to converge at <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\mathcal {O}(1/\sqrt{t})$</tex-math></inline-formula> prior to our work. In the second one, we use the extrapolation technique to accelerate the convergence of DDA. Compared to existing accelerated algorithms, where typically two different variables are exchanged among agents at each time, the proposed algorithm only seeks consensus on local gradients. Then, the extrapolation is performed based on two sequences of primal variables, which are determined by the accumulations of gradients at two consecutive time instants, respectively. The algorithm is proved to converge at <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\mathcal {O}(1)\left(\frac{1}{t^{2}}+\frac{1}{t(1-\beta)^{2}}\right)$</tex-math></inline-formula> , where <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\beta$</tex-math></inline-formula> denotes the second largest singular value of the mixing matrix. We remark that the condition for the algorithmic parameter to guarantee convergence does not rely on the spectrum of the mixing matrix, making itself easy to satisfy in practice. Finally, numerical results are presented to demonstrate the efficiency of the proposed methods.

Marketing Concepts of New Housing Projects in Surabaya during Pandemic based on Property Agent Perspective

The increasing demand and needs of consumers in property transactions makes the need for property agents even greater. In a pandemic condition that limits direct interaction, many housing developers have difficulty selling their housing products due to limited interaction with potential consumers and the lack of up-to-date information on housing products that consumers are interested in during this pandemic. The purpose of this study was to analyze the factors and sequence of criteria that became the basis for selecting new housing during the pandemic based on the perspective of a property agent.This study uses qualitative and quantitative research using interview data and processed into numerical data using product, price, promotion, place, and person variables. The method used is the Delphi method, AHP and Content Analysis, with the Delphi method useful for determining criteria and indicators that are validated from experts and AHP functioning as a weight for the validated criteria. While Content Analysis is used to formulate concepts.The results of this study obtained a sequence of variables that affect Product, (2) Place, (3) Price, (4) Promotion, (5) People.

Neuroborreliosis and Post-Treatment Lyme Disease Syndrome: Focus on Children.

Neuroborreliosis is a form of Lyme Borreliosis (LB) that affects various structures of the central and peripheral nervous system. Although most cases of LB can be cured with a course of antibiotics, some children can present prolonged symptoms, which may constitute post-treatment Lyme disease syndrome (PTLDS). The aim of our analysis was the long-term observation of children with NB and the determination of their risk of PTLDS. The clinical observation was supplemented by a laboratory study based on the assessment of the dynamics of anti-VlsE (variable major protein-like sequence, expressed) IgG antibodies in children with NB after antibiotic therapy. The prospective survey based on 40 children presented 1-2 forms of NB. The control group consisted of 36 patients with analogical symptoms for whom LB was excluded. Our long-term observation showed a low risk of developing long-term complications in children who received antibiotic therapy in accordance with the recommendations. The concentration of anti-VlsE IgG demonstrates a statistical significance for differences between the control and the study groups for each measurement period. Higher values of anti-VlsE IgG were observed in the study group, and the concentration decreased from the first measurement period to the next. The article emphasizes the importance of the long-term follow-up of children with neuroborreliosis.

Examining the causal model between socially shared regulation of motivation, engagement, and creative performance

The purpose of this study was to hypothesize and validate a path model with the following sequence of variables: three regulation modes of intrinsic motivation, three aspects of engagement, and creative performance. In a traditional lecture class with a large number of students, opportunities for pair work and individual's writing were incorporated to promote student's creative and deeper thinking. Participants were 325 undergraduate students who were asked to complete a questionnaire on self-regulation, co-regulation, and socially shared regulation of intrinsic motivation before the class, and behavioral, emotional, and cognitive engagement after the class. The results of path analysis using structural equation modeling showed that self-regulation of motivation positively influenced behavioral, emotional, and cognitive engagement. Additionally, co-regulation of motivation positively affected emotional and cognitive engagement. Finally, behavioral engagement positively influenced creative performance. Based on the findings of the path model, the desirable direction of educational practice in the future was discussed.

Deferred Nörlund statistical convergence in probability, mean and distribution for sequences of random variables

Deferred Nörlund statistical convergence in probability, mean and distribution for sequences of random variables

CLL stereotyped B-cell receptor immunoglobulin sequences are recurrent in the B-cell repertoire of healthy individuals: Apparent lack of central and early peripheral tolerance censoring.

The leukemic cells of patients with chronic lymphocytic leukemia (CLL) are often unique, expressing remarkably similar IGHV-IGHD-IGHJ gene rearrangements, "stereotyped BCRs". The B-cell receptors (BCRs) on CLL cells are also distinctive in often deriving from autoreactive B lymphocytes, leading to the assumption of a defect in immune tolerance. Using bulk and single-cell immunoglobulin heavy and light chain variable domain sequencing, we enumerated CLL stereotype-like IGHV-IGHD-IGHJ sequences (CLL-SLS) in B cells from cord blood (CB) and adult peripheral blood (PBMC) and bone marrow (BM of healthy donors. CLL-SLS were found at similar frequencies among CB, BM, and PBMC, suggesting that age does not influence CLL-SLS levels. Moreover, the frequencies of CLL-SLS did not differ among B lymphocytes in the BM at early stages of development, and only re-circulating marginal zone B cells contained significantly higher CLL-SLS frequencies than other mature B-cell subpopulations. Although we identified CLL-SLS corresponding to most of the CLL major stereotyped subsets, CLL-SLS frequencies did not correlate with those found in patients. Interestingly, in CB samples, half of the CLL-SLS identified were attributed to two IGHV-mutated subsets. We also found satellite CLL-SLS among the same normal samples, and they were also enriched in naïve B cells but unexpectedly, these were ~10-fold higher than standard CLL-SLS. In general, IGHV-mutated CLL-SLS subsets were enriched among antigen-experienced B-cell subpopulations, and IGHV-unmutated CLL-SLS were found mostly in antigen-inexperienced B cells. Nevertheless, CLL-SLS with an IGHV-mutation status matching that of CLL clones varied among the normal B-cell subpopulations, suggesting that specific CLL-SLS could originate from distinct subpopulations of normal B cells. Lastly, using single-cell DNA sequencing, we identified paired IGH and IGL rearrangements in normal B lymphocytes resembling those of stereotyped BCRs in CLL, although some differed from those in patients based on IG isotype or somatic mutation. CLL-SLS are present in normal B-lymphocyte populations at all stages of development. Thus, despite their autoreactive profile they are not deleted by central tolerance mechanisms, possibly because the level of autoreactivity is not registered as dangerous by deletion mechanisms or because editing of L-chain variable genes occurred which our experimental approach could not identify.

Test-retest repeatability of ADC in prostate using the multi b-Value VERDICT acquisition

PurposeVERDICT (Vascular, Extracellular, Restricted Diffusion for Cytometry in Tumours) MRI is a multi b-value, variable diffusion time DWI sequence that allows generation of ADC maps from different b-value and diffusion time combinations. The aim was to assess precision of prostate ADC measurements from varying b-value combinations using VERDICT and determine which protocol provides the most repeatable ADC. Materials and MethodsForty-one men (median age: 67.7 years) from a prior prospective VERDICT study (April 2016–October 2017) were analysed retrospectively. Men who were suspected of prostate cancer and scanned twice using VERDICT were included. ADC maps were formed using 5b-value combinations and the within-subject standard deviations (wSD) were calculated per ADC map. Three anatomical locations were analysed per subject: normal TZ (transition zone), normal PZ (peripheral zone), and index lesions. Repeated measures ANOVAs showed which b-value range had the lowest wSD, Spearman correlation and generalized linear model regression analysis determined whether wSD was related to ADC magnitude and ROI size. ResultsThe mean lesion ADC for b0b1500 had the lowest wSD in most zones (0.18–0.58x10-4 mm2/s). The wSD was unaffected by ADC magnitude (Lesion: p = 0.064, TZ: p = 0.368, PZ: p = 0.072) and lesion Likert score (p = 0.95). wSD showed a decrease with ROI size pooled over zones (p = 0.019, adjusted regression coefficient = -1.6x10-3, larger ROIs for TZ versus PZ versus lesions). ADC maps formed with a maximum b-value of 500 s/mm2 had the largest wSDs (1.90–10.24x10-4 mm2/s). ConclusionADC maps generated from b0b1500 have better repeatability in normal TZ, normal PZ, and index lesions.

IMPLEMENTATION OF HEALTH-SAVING TECHNOLOGIES IN SWIMMING LESSONS

The article defines the peculiarities of the implementation of health-saving technologies in swimming lessons in higher education institutions. Modern health-saving technologies used in the field of physical education of student youth are disclosed. The current state of health and physical fitness of student swimmers was analyzed, a set of problems related to the development of sports and mass swimming in Ukraine was identified, in particular, regarding the training of the country's population and its conscript contingents in the ability to swim. The role and place of swimming in the system of physical fitness of student youth and ensuring their adherence to the basics of a healthy lifestyle in the context of the implementation of health-saving technologies are determined.&#x0D; Swimming training is carried out in the form of practical classes, which take up one training pair (2 academic hours).&#x0D; Methodical planning and structuring of the training process, which includes the logical and systematic ordering of several training variables (intensity, volume, frequency, recovery period and exercises) in an integrative way aimed at optimizing specific performance results at predetermined points in time.&#x0D; The main characteristics of such approaches are systematic variations in the content of the intensity of classes and the amount of workload throughout the entire program.&#x0D; However, this should not be seen as a simple strategy of training variation, but as an appropriate sequence and integration of different training variables, including not only volume and intensity, but also frequency, recovery periods, density and selection of exercises.&#x0D; The duration of each session may increase. The total training contribution of strength work on land is a product of intensity, frequency and duration of training. The general goal is the uniform development of all muscle groups.&#x0D; Land exercises are used to train muscular strength and endurance, and they complement the improvements in muscular endurance that result from swimming training. The exercise program should be simple (that is, two exercises for each of the main muscle groups) so that it corresponds to the total training time for one session.

AlphaFlow: autonomous discovery and optimization of multi-step chemistry using a self-driven fluidic lab guided by reinforcement learning

Closed-loop, autonomous experimentation enables accelerated and material-efficient exploration of large reaction spaces without the need for user intervention. However, autonomous exploration of advanced materials with complex, multi-step processes and data sparse environments remains a challenge. In this work, we present AlphaFlow, a self-driven fluidic lab capable of autonomous discovery of complex multi-step chemistries. AlphaFlow uses reinforcement learning integrated with a modular microdroplet reactor capable of performing reaction steps with variable sequence, phase separation, washing, and continuous in-situ spectral monitoring. To demonstrate the power of reinforcement learning toward high dimensionality multi-step chemistries, we use AlphaFlow to discover and optimize synthetic routes for shell-growth of core-shell semiconductor nanoparticles, inspired by colloidal atomic layer deposition (cALD). Without prior knowledge of conventional cALD parameters, AlphaFlow successfully identified and optimized a novel multi-step reaction route, with up to 40 parameters, that outperformed conventional sequences. Through this work, we demonstrate the capabilities of closed-loop, reinforcement learning-guided systems in exploring and solving challenges in multi-step nanoparticle syntheses, while relying solely on in-house generated data from a miniaturized microfluidic platform. Further application of AlphaFlow in multi-step chemistries beyond cALD can lead to accelerated fundamental knowledge generation as well as synthetic route discoveries and optimization.

Adaptive control for downhole nuclear magnetic resonance excitation

Nuclear magnetic resonance (NMR) measurements are performed with the pulse sequence and acquisition parameters set by the operator, which cannot be adjusted in real time according to sample characteristics. In one acquisition cycle, usually thousands of high-power pulses are transmitted and thousands of echo points are acquired. The power consumption cause the RF amplifier to overheat, and large amounts of acquired data may be invalid. Therefore, the optimization of excitation and acquisition processes is necessary to improve measurement efficiency. We explore a scheme for the real-time measurement of the samples by adaptively regulating the pulse sequence, which adapts the variable TE pulse sequence as the reconnaissance mode. The appropriate pulse sequence and reasonable parameters (NE, TE) can be selected according to the relaxation characteristics of the samples.This adaptive control strategy has great significance in guiding both dynamic and static measurements, and it is especially suitable for occasions where low magnetic field gradients and diffusion terms can be ignored. We also design a test circuit for adaptive control, which has the function of automatic parameter adjustment. By adjusting parameters such as the number of refocusing pulses, echo spacing, etc., the effective measurement of the samples can be achieved in practice.