Search

ObjectiveThe objective of this study was to investigate the spatio-temporal distribution and epidemiological characteristics of hepatitis B in 96 districts and counties of Xinjiang and to give useful information for hepatitis B prevention and treatment.MethodsBased on the incidence data of hepatitis B in 96 districts and counties of Xinjiang from 2006 to 2019, the global trend analysis method was used to characterize the spatial variability of the disease, and the spatial autocorrelation and spatio-temporal aggregation analysis were used to explore the spatial clustering of hepatitis B and to identify high-risk areas and periods. The Integrated Nested Laplace Approximation (INLA)-based spatial age-period-cohort model was established to further explore the influence of age, period, birth queue effect, and spatial distribution on the incidence risk of hepatitis B, and sum-to-zero constraint was adopted to avoid the issue of model unrecognition.ResultsThe risk of hepatitis B in Xinjiang is increasing from west to east and from north to south, with spatial heterogeneity and spatio-temporal scanning statistics yielding five clustering areas. The spatial age-period-cohort model showed two peaks in the average risk of hepatitis B, at [25,30) years old and [50,55) years old, respectively. The mean risk of hepatitis B incidence fluctuated up and down around 1 with time, and the average risk of disease by birth cohort displayed an increasing-decreasing-stabilizing trend. Taking age, period, and cohort effect into consideration, it was found that the areas with a high risk of hepatitis B are Tianshan District, Xinshi District, Shuimogou District, Changji City, Aksu City, Kashi City, Korla City, Qiemo County and Yopurga County in Xinjiang. According to the spatio-temporal effect item, it was found that there are unobserved variables affecting the incidence of hepatitis B in some districts and counties of Xinjiang.ConclusionThe spatio-temporal characteristics of hepatitis B and the high-risk population needed to be taken into attention. It is suggested that the relevant disease prevention and control centers should strengthen the prevention and control of hepatitis B among young people while paying attention to middle-aged and older adult people, and strengthening the prevention and monitoring of high-risk areas.

Macrophages are the core of the pathophysiology of rheumatoid arthritis (RA). They participate in specific and non-specific immunological responses, have phagocytosis, chemotaxis and immune regulatory functions, and are involved in the onset and progression of RA. In recent years, research on the pathophysiology of RA has focused on the polarization and functions of classically activated M1 and selectively activated M2 macrophage subtypes. M1 macrophages release different proinflammatory cytokines, thus driving the chronic proinflammatory, tissue destruction and pain response in RA. M2 macrophages play an anti-inflammatory role. Because of the important role of monocyte-macrophage in RA, drug research targeting monocyte-macrophage can bring us more hope for treatment of RA. This study reviewed the characteristics, plasticity, molecular activation mechanism and relationship of RA with mononuclear macrophages, as well as the transformative potential of macrophages in developing new therapeutic drugs for clinical practice.

It is very difficult to forecast the production rate of oil wells as the output of a single well is sensitive to various uncertain factors, which implicitly or explicitly show the influence of the static, temporal and spatial properties on the oil well production. In this study, a novel machine learning model is constructed to fuse the static geological information, dynamic well production history, and spatial information of the adjacent water injection wells. There are 3 basic modules in this stacking model, which are regarded as the encoders to extract the features from different types of data. One is Multi-Layer Perceptron, which is to analyze the static geological properties of the reservoir that might influence the well production rate. The other two are both LSTMs, which have the input in the form of two matrices rather than vectors, standing for the temporal and the spatial information of the target well. The difference of the two modules is that in the spatial information processing module we take into consideration the time delay of water flooding response, from the injection well to the target well. In addition, we use Symbolic Transfer Entropy to prove the superiorities of the stacking model from the perspective of Causality Discovery. It is proved theoretically and practically that the presented model can make full use of the model structure to integrate the characteristics of the data and the experts' knowledge into the process of machine learning, greatly improving the accuracy and generalization ability of prediction.

We consider infinite horizon Markov decision processes (MDPs) with fast-slow structure, meaning that certain parts of the state space move "fast" (and in a sense, are more influential) while other parts transition more "slowly." Such structure is common in real-world problems where sequential decisions need to be made at high frequencies, yet information that varies at a slower timescale also influences the optimal policy. Examples include: (1) service allocation for a multi-class queue with (slowly varying) stochastic costs, (2) a restless multi-armed bandit with an environmental state, and (3) energy demand response, where both day-ahead and real-time prices play a role in the firm's revenue. Models that fully capture these problems often result in MDPs with large state spaces and large effective time horizons (due to frequent decisions), rendering them computationally intractable. We propose an approximate dynamic programming algorithmic framework based on the idea of "freezing" the slow states, solving a set of simpler finite-horizon MDPs (the lower-level MDPs), and applying value iteration (VI) to an auxiliary MDP that transitions on a slower timescale (the upper-level MDP). We also extend the technique to a function approximation setting, where a feature-based linear architecture is used. On the theoretical side, we analyze the regret incurred by each variant of our frozen-state approach. Finally, we give empirical evidence that the frozen-state approach generates effective policies using just a fraction of the computational cost, while illustrating that simply omitting slow states from the decision modeling is often not a viable heuristic., 69 pages, 9 figures

We propose a general framework for decoding quantum error-correcting codes with generative modeling. The model utilizes autoregressive neural networks, specifically Transformers, to learn the joint probability of logical operators and syndromes. This training is in an unsupervised way, without the need for labeled training data, and is thus referred to as pre-training. After the pre-training, the model can efficiently compute the likelihood of logical operators for any given syndrome, using maximum likelihood decoding. It can directly generate the most-likely logical operators with computational complexity $\mathcal O(2k)$ in the number of logical qubits $k$, which is significantly better than the conventional maximum likelihood decoding algorithms that require $\mathcal O(4^k)$ computation. Based on the pre-trained model, we further propose refinement to achieve more accurately the likelihood of logical operators for a given syndrome by directly sampling the stabilizer operators. We perform numerical experiments on stabilizer codes with small code distances, using both depolarizing error models and error models with correlated noise. The results show that our approach provides significantly better decoding accuracy than the minimum weight perfect matching and belief-propagation-based algorithms. Our framework is general and can be applied to any error model and quantum codes with different topologies such as surface codes and quantum LDPC codes. Furthermore, it leverages the parallelization capabilities of GPUs, enabling simultaneous decoding of a large number of syndromes. Our approach sheds light on the efficient and accurate decoding of quantum error-correcting codes using generative artificial intelligence and modern computational power., Comment: Comments are welcome

When studying interacting systems, computing their statistical properties is a fundamental problem in various fields such as physics, applied mathematics, and machine learning. However, this task can be quite challenging due to the exponential growth of the state space as the system size increases. Many standard methods have significant weaknesses. For instance, message-passing algorithms can be inaccurate and even fail to converge due to short loops. At the same time, tensor network methods can have exponential computational complexity in large graphs due to long loops. This work proposes a new method called ``tensor network message passing.'' This approach allows us to compute local observables like marginal probabilities and correlations by combining the strengths of tensor networks in contracting small sub-graphs with many short loops and the strengths of message-passing methods in globally sparse graphs, thus addressing the crucial weaknesses of both approaches. Our algorithm is exact for systems that are globally tree-like and locally dense-connected when the dense local graphs have limited treewidth. We have conducted numerical experiments on synthetic and real-world graphs to compute magnetizations of Ising models and spin glasses, to demonstrate the superiority of our approach over standard belief propagation and the recently proposed loopy message-passing algorithm. In addition, we discuss the potential applications of our method in inference problems in networks, combinatorial optimization problems, and decoding problems in quantum error correction., Comment: Comments are welcome!

IntroductionCleft palate (CP) is one of the most common congenital birth defects in the craniofacial region, retinoic acid (RA) gavage is the most common method for inducing cleft palate model. Although several mechanisms have been proposed to illuminate RA-induced cleft palate during embryonic development, these findings are far from enough. Many efforts remain to be devoted to studying the etiology and pathogenesis of cleft palate. Recent research is gradually shifting the focus to the effect of retinoic acid on gut microbiota. However, few reports focus on the relationship between the occurrence of CP in embryos and gut microbiota.MethodsIn our research, we used RA to induce cleft palate model for E10.5 the feces of 5 RA-treated pregnant mice and 5 control pregnant mice were respectively metagenomics analysis.ResultsCompared with the control group, Lactobacillus in the gut microbiome the RA group was significantly increased. GO, KEGG and CAZy analysis of differentially unigenes demonstrated the most abundant metabolic pathway in different groups, lipopolysaccharide biosynthesis, and histidine metabolism.DiscussionOur findings indicated that changes in the maternal gut microbiome palatal development, which might be related to changes in Lactobacillus and These results provide a new direction in the pathogenesis of CP induced by RA.

Calmodulin-binding transcription activators (CAMTAs) are evolutionarily conserved transcription factors and have multi-functions in plant development and stress response. However, identification and functional analysis of tea plant (Camellia sinensis) CAMTA genes (CsCAMTAs) are still lacking. Here, five CsCAMTAs were identified from tea plant genomic database. Their gene structures were similar except CsCAMTA2, and protein domains were conserved. Phylogenetic relationship classified the CsCAMTAs into three groups, CsCAMTA2 was in group I, and CsCAMTA1, 3 and CsCAMTA4, 5 were, respectively, in groups II and III. Analysis showed that stress and phytohormone response-related cis-elements were distributed in the promoters of CsCAMTA genes. Expression analysis showed that CsCAMTAs were differentially expressed in different organs and under various stress treatments of tea plants. Three-hundred and four hundred-one positive co-expressed genes of CsCAMTAs were identified under cold and drought, respectively. CsCAMTAs and their co-expressed genes constituted five independent co-expression networks. KEGG enrichment analysis of CsCAMTAs and the co-expressed genes revealed that hormone regulation, transcriptional regulation, and protein processing-related pathways were enriched under cold treatment, while pathways like hormone metabolism, lipid metabolism, and carbon metabolism were enriched under drought treatment. Protein interaction network analysis suggested that CsCAMTAs could bind (G/A/C)CGCG(C/G/T) or (A/C)CGTGT cis element in the target gene promoters, and transcriptional regulation might be the main way of CsCAMTA-mediated functional regulation. The study establishes a foundation for further function studies of CsCAMTA genes in stress response.

Eigen mode selection ought to be a practical issue in some real game systems, as it is a practical issue in the dynamics behaviour of a building, bridge, or molecular, because of the mathematical similarity in theory. However, its reality and accuracy have not been known in real games. We design a 5-strategy game which, in the replicator dynamics theory, is predicted to exist two eigen modes. Further, in behaviour game theory, the game is predicted that the mode selection should depends on the game parameter. We conduct human subject game experiments by controlling the parameter. The data confirm that, the predictions on the mode existence as well as the mode selection are significantly supported. This finding suggests that, like the equilibrium selection concept in classical game theory, eigen mode selection is an issue in game dynamics theory., game dynamics theory, eigen mode, human subject, experimental economics, behavior game theory, eigenvector, eigencycle

Game dynamics theory, as a field of science, the consistency of theory and experiment is essential. In the past 10 years, important progress has been made in the merging of the theory and experiment in this field, in which dynamics cycle is the presentation. However, the merging works have not got rid of the constraints of Euclidean two-dimensional cycle so far. This paper uses a classic four-strategy game to study the dynamic structure (non-Euclidean superplane cycle). The consistency is in significant between the three ways: (1) the analytical results from evolutionary dynamics equations, (2) agent-based simulation results from learning models and (3) laboratory results from human subjects game experiments. The consistency suggests that, game dynamic structure could be quantitatively predictable, observable and controllable in general., game theory; laboratory game experiment; eigenvector; dynamics system theory

With the process of Vietnam's reform and opening up, Vietnam's economy has made remarkable achievements. Vietnam's manufacturing industry, taking advantage of the new liberal policies, has also achieved rapid development. Sufficient and cheap labor force, advantageous geographical location, preferential foreign investment policies, and friendly international trade environment with the technological upgrading of manufacturing industry are all the advantages of Vietnam in attracting manufacturing industry. However, there are structural difficulties in the Vietnam's economy. Vietnam's economy is highly dependent on foreign trade and foreign investment, and its trade commodities are mainly assembly and processing with low added value. Compared with China, Vietnam also has obvious disadvantages in the scale of domestic market and supply chain. To some extent, Vietnam's manufacturing industry is integrated into China's supply chain network. Keywords: manufacturing, foreign trade, Vietnam's economy, comparative advantage, supply chain

G proteins function directly in cold tolerance of plants. However, the framework of the Gα subunit in regulating cold tolerance remains to be explored. Here, we used protein interaction techniques to elucidate cold-related pathways regulated by CsGPA1. Suppression of

Based on the number of 519 tunnel risk incidents that occurred in the two tunnels of the Qifu Tunnel and the Zhongcun Tunnel during the period from 2018 to 2019 of Guangming Expressway, this paper studies the temporal and spatial characteristics and the risk distribution of the sunken continuous tunnel during the operation period of the expressway. Type of risk event. The results show that January, May, June and September of the year, as well as 14:00–16:00 and 16:00–18:00 during the day are periods of high tunnel risk; at the entrance section of continuous tunnels, Compared with other locations, the number of risk events in the transition section and the open section with sudden environmental changes and gradient changes is more; the types of risk events include safety hazards, roadblocks, vehicle failures, rear-end collisions, and equipment failures. The main types are vehicle failures. There are certain differences in the east-west direction. There are more vehicle breakdowns in the east-bound direction, more roadblocks in the west-bound direction, and more rear-end collisions in the east-bound direction. The main types of risk events are cars and trucks. Both cars and trucks have major risk event types. It is a vehicle failure. In rear-end collisions, small cars account for 65% of the risk models; risk identification methods include gun patrol discovery, road administration reporting, etc., of which gun patrol discovery is the most important identification method, accounting for 65% of the total. Through the analysis of the risk event characteristics of the sunken continuous tunnel of the expressway, it provides reference opinions for perfecting the research deficiencies in related fields in our country.

Molecularly imprinted polymers (MIPs) are chemically synthesized materials mimicking the recognition of antibodies towards antigens., Molecularly imprinted polymers (MIPs) are chemically synthesized materials mimicking the recognition of antibodies towards antigens. Epitope imprinting has been an effective strategy, making imprinting of proteins flexible to a great extent. However, so far there is apparently a lack of facile and versatile epitope imprinting approaches. Herein, we present a new method called controllable oriented surface imprinting of boronate affinity-anchored epitopes. In this method, a C-terminus nonapeptide epitope was glycated and anchored as a template onto a boronic acid-functionalized substrate, followed by controllable oriented surface imprinting via the polycondensation of multiple silylating reagents containing functionalities capable of interacting with the epitope. The developed imprinting approach allowed for precise control of the thickness of the imprinting layer through adjusting the imprinting time, generating excellent binding properties. This method was verified to be versatile and efficient. Thus, it could greatly facilitate the preparation of MIPs for specific recognition of proteins and peptides.

Deep learning models in large-scale machine learning systems are often continuously trained with enormous data from production environments. The sheer volume of streaming training data poses a significant challenge to real-time training subsystems and ad-hoc sampling is the standard practice. Our key insight is that these deployed ML systems continuously perform forward passes on data instances during inference, but ad-hoc sampling does not take advantage of this substantial computational effort. Therefore, we propose to record a constant amount of information per instance from these forward passes. The extra information measurably improves the selection of which data instances should participate in forward and backward passes. A novel optimization framework is proposed to analyze this problem and we provide an efficient approximation algorithm under the framework of Mini-batch gradient descent as a practical solution. We also demonstrate the effectiveness of our framework and algorithm on several large-scale classification and regression tasks, when compared with competitive baselines widely used in industry., 13 pages

Yijia Wang,1,* Chunze Zhang,2,* Shaobin Hou,3 Xiaojing Wu,1 Jun Liu,4 Xuehua Wan5 1Laboratory of Oncologic Molecular Medicine, Tianjin Union Medical Center, Nankai University, Tianjin, People’s Republic of China; 2Department of Colorectal Surgery, Tianjin Union Medical Center, Nankai University, Tianjin, People’s Republic of China; 3Advanced Studies in Genomics, Proteomics, and Bioinformatics, University of Hawaii at Manoa, Honolulu, HI, USA; 4Department of Radiology, Tianjin Union Medical Center, Nankai University, Tianjin, People’s Republic of China; 5TEDA Institute of Biological Sciences and Biotechnology, Nankai University, TEDA, Tianjin, People’s Republic of China*These authors contributed equally to this workCorrespondence: Jun LiuTianjin Union Medical Center, No. 190, Jieyuan Road, Hongqiao District, Tianjin City, People’s Republic of ChinaEmail junliu_sci@163.comXuehua WanNankai University, No. 94, Weijin Road, Nankai District, Tianjin City, People’s Republic of ChinaEmail xuehua.wan@hotmail.comIntroduction: Besides genetic and epigenetic alterations that lead to carcinogenesis and development of colorectal cancer (CRC), intestinal microbiomes are recently recognized to play a critical role in CRC progression. The abundant species associated with human CRC have been proposed for their roles in promoting tumorigenesis. However, a recent “driver-passenger” model suggests that these CRC-associated species with high relative abundances may be passenger bacteria that take advantage of the tumor environment instead of initiating CRC, whereas the driver species that initiate CRC have been replaced by passenger bacteria due to the alteration of the intestinal niche.Methods: Here, to reveal potential driver and passenger bacteria during CRC progression, we compare the gut mucosal microbiomes of 75 triplet-paired CRC samples collected from on-tumor site, adjacent-tumor site, and off-tumor site, and 26 healthy controls.Results: Our analyses revealed potential driver bacteria in four genera and two families, and potential passenger bacteria in 14 genera or families. Bacillus, Bradyrhizobium, Methylobacterium, Streptomyces, Intrasporangiaceae and Sinobacteraceae were predicted to be potential driver bacteria. Moreover, 14 potential passenger bacteria were identified and divided into five groups. Group I passenger bacteria contain Fusobacterium, Campylobacter, Streptococcus, Schwartzia, and Parvimonas. Group II passenger bacteria contain Dethiosulfatibacter, Selenomonas, Peptostreptococus, Leptotrichia. Group III passenger bacteria contain Granulicatella. Group IV passenger bacteria contain Shewanella, Mogibacterium, and Eikenella. Group V passenger bacteria contain Anaerococus. Co-occurrence network analysis reveals a low correlation relationship between driver and passenger bacteria in CRC patients compared with healthy controls.Discussion: These driver and passenger species may serve as bio-marker species for screening cohorts with high risk to initiate CRC or patients with CRC, respectively. Further functional studies will help understand the roles of driver and passenger bacteria in CRC initiation and development.Keywords: colorectal cancer, driver-passenger model, microbiota

We study the problem of learning the objective functions or constraints of a multiobjective decision making model, based on a set of sequentially arrived decisions. In particular, these decisions might not be exact and possibly carry measurement noise or are generated with the bounded rationality of decision makers. In this paper, we propose a general online learning framework to deal with this learning problem using inverse multiobjective optimization. More precisely, we develop two online learning algorithms with implicit update rules which can handle noisy data. Numerical results show that both algorithms can learn the parameters with great accuracy and are robust to noise., Comment: 17 pages

Policy optimization in unknown, sparse-reward environments with expensive and limited interactions is challenging, and poses a need for effective exploration. Motivated by complex navigation tasks that require real-world training (when cheap simulators are not available), we consider an agent that faces an unknown distribution of environments and must decide on an exploration strategy, through a series of training environments, that can benefit policy learning in a test environment drawn from the environment distribution. Most existing approaches focus on fixed exploration strategies, while the few that view exploration as a meta-optimization problem tend to ignore the need for cost-efficient exploration. We propose a cost-aware Bayesian optimization approach that efficiently searches over a class of dynamic subgoal-based exploration strategies. The algorithm adjusts a variety of levers -- the locations of the subgoals, the length of each episode, and the number of replications per trial -- in order to overcome the challenges of sparse rewards, expensive interactions, and noise. Our experimental evaluation demonstrates that, when averaged across problem domains, the proposed algorithm outperforms the meta-learning algorithm MAML by 19%, the hyperparameter tuning method Hyperband by 23%, BO techniques EI and LCB by 24% and 22%, respectively. We also provide a theoretical foundation and prove that the method asymptotically identifies a near-optimal subgoal design from the search space., Comment: Presented at TARL, ICLR 2019 workshop

There are successful and failure cases of PPP financing model projects in China. Combined with the respective interests of the two main participants of the PPP project, the government and social capital, the application analysis of the PPP model in China’s infrastructure sector put forward the view that demand analysis plays a key role in the success or failure of the PPP project. Based on this, combined with the characteristics of infrastructure, the characteristics and production of infrastructure rigid demand and elastic demand are briefly analyzed. The typical basic resources of water resources are taken as examples to analyze the rigid demand and elastic demand of water resources,and how to analyze the demand analysis of PPP projects, and explore them in cross-regional water markets.

With the development of the EV market, the demand for charging facilities is growing rapidly. The rapid increase in Electric Vehicle and different market factors bring challenges to the prediction of the penetration rate of EV number. The estimates of the uptake rate of EVs for light passenger use vary widely with some scenarios gradual and others aggressive. And there have been many effects on EV penetration rate from incentives, tax breaks, and market price. Given this background, this research is devoted to addressing a stochastic joint planning framework for both EV charging system and distribution network where the EV behaviours in both transportation network and electrical system are considered. And the planning issue is formulated as a multi-objective model with both the capital investment cost and service convenience optimized. The optimal planning of EV charging system in the urban area is the target geographical planning area in this work where the service radius and driving distance is relatively limited. The mathematical modelling of EV driving and charging behaviour in the urban area is developed.

Knowing the reflection of game theory and ethics, we develop a mathematical representation to bridge the gap between the concepts in moral philosophy (e.g., Kantian and Utilitarian) and AI ethics industry technology standard (e.g., IEEE P7000 standard series for Ethical AI). As an application, we demonstrate how human value can be obtained from the experimental game theory (e.g., trust game experiment) so as to build an ethical AI. Moreover, an approach to test the ethics (rightness or wrongness) of a given AI algorithm by using an iterated Prisoner's Dilemma Game experiment is discussed as an example. Compared with existing mathematical frameworks and testing method on AI ethics technology, the advantages of the proposed approach are analyzed., 6 pages, 8 figures

As the stress is one of the main factors affecting the safe operation of the pipeline in the gas pipeline tunnel crossing project, in order to ensure the safe operation of the pipeline, it is necessary to research the stress conditions of the gas pipelines in the deviated well. In this study, we discuss on these two aspects of the gas pipelines in the deviated well: the necessity of adding buttresses and the distance between buttresses. We do the numerical simulation of China Shaanxi-Beijing III gas pipeline in the Eighth Fort Yellow River tunnel areas using stress analysis software CAESAR II; exploring its secure and economical buttresses distance. Our research shows that the pipeline can be safely run after adding buttresses and it is more reasonable to set buttresses distance of 20 m for the numerical example. Our research provides a reference for the design of the buttresses distance of gas pipelines in the deviated well and has some engineering value.

The sensitivity and dynamic range are main performance parameters of a Surface Plasmon Resonance(SPR) biosensor with phase modulation.In this paper,the main effect factors on sensitivity and dynamic range of the sensor were analyzed,and the influence of nonlinear change of phase difference on the results of real-time monitoring biological reaction was analyzed.A high resolution SPR system based on a Mach-Zehnder configuration was set up.The phase difference curves of a series gold membrane thicknesses and incident angles were simulated by matlab software,and the effect of several factors on the resolution and dynamic range were evaluated.In addition,the real-time monitoring of binding reaction between Bovine Serum Albumin(BSA) and BSA antibodies was also demonstrated.Obtainecl results show that the influence of membrane thickness is significant and nonlinear,but the incident angle shows little effect on the resolution and dynamic range.The influence of narrow dynamic range on biology reaction measurement can be minimized by optimizing some parameters,such as membrane thickness,incident angle and reactants concentration.Experimental results show that the sensitivity and dynamic range can be optimized by adjusting gold membrane thicknesses.This paper analyzed several influence factors of the sensitivity and dynamic range of phase modulation SPR biosensors.For a light source with 633 nm wavelength,the most optimal membrane thickness is 48 nm when the reaction between BSA and its antibody is measured.In this situation,the dynamic range is 0.013 6RIU and the sensitivity is 6.67×10-7RIU/0.01°.

Asymmetric cell division is critical for generating cell diversity in low eukaryotic organisms. We previously have reported that polyploid giant cancer cells (PGCCs) induced by cobalt chloride demonstrate the ability to use an evolutionarily conserved process for renewal and fast reproduction, which is normally confined to simpler organisms. The budding yeast, Saccharomyces cerevisiae, which reproduces by asymmetric cell division, has long been a model for asymmetric cell division studies. PGCCs produce daughter cells asymmetrically in a manner similar to yeast, in that both use budding for cell polarization and cytokinesis. Here, we review the results of recent studies and discuss the similarities in the budding process between yeast and PGCCs.