PUFFFIN: an ultra-bright, customisable, single-plasmid system for labelling cell neighbourhoods.

Cell communication coordinates developmental processes, maintains homeostasis, and contributes to disease. Therefore, understanding the relationship between cells in a shared environment is crucial. Here we introduce Positive Ultra-bright Fluorescent Fusion For Identifying Neighbours (PUFFFIN), a cell neighbour-labelling system based upon secretion and uptake of positively supercharged fluorescent protein s36GFP. We fused s36GFP to mNeonGreen or to a HaloTag, facilitating ultra-bright, sensitive, colour-of-choice labelling. Secretor cells transfer PUFFFIN to neighbours while retaining nuclear mCherry, making identification, isolation, and investigation of live neighbours straightforward. PUFFFIN can be delivered to cells, tissues, or embryos on a customisable single-plasmid construct composed of interchangeable components with the option to incorporate any transgene. This versatility enables the manipulation of cell properties, while simultaneously labelling surrounding cells, in cell culture or in vivo. We use PUFFFIN to ask whether pluripotent cells adjust the pace of differentiation to synchronise with their neighbours during exit from naïve pluripotency. PUFFFIN offers a simple, sensitive, customisable approach to profile non-cell-autonomous responses to natural or induced changes in cell identity or behaviour.

Enabling neighbour labelling: using synthetic biology to explore how cells influence their neighbours.

Cell-cell interactions are central to development, but exploring how a change in any given cell relates to changes in the neighbour of that cell can be technically challenging. Here, we review recent developments in synthetic biology and image analysis that are helping overcome this problem. We highlight the opportunities presented by these advances and discuss opportunities and limitations in applying them to developmental model systems.

Mechanical stress combines with planar polarised patterning during metaphase to orient embryonic epithelial cell divisions.

The planar orientation of cell division (OCD) is important for epithelial morphogenesis and homeostasis. Here, we ask how mechanics and antero-posterior (AP) patterning combine to influence the first divisions after gastrulation in the Drosophila embryonic epithelium. We analyse hundreds of cell divisions and show that stress anisotropy, notably from compressive forces, can reorient division directly in metaphase. Stress anisotropy influences the OCD by imposing metaphase cell elongation, despite mitotic rounding, and overrides interphase cell elongation. In strongly elongated cells, the mitotic spindle adapts its length to, and hence its orientation is constrained by, the cell long axis. Alongside mechanical cues, we find a tissue-wide bias of the mitotic spindle orientation towards AP-patterned planar polarised Myosin-II. This spindle bias is lost in an AP-patterning mutant. Thus, a patterning-induced mitotic spindle orientation bias overrides mechanical cues in mildly elongated cells, whereas in strongly elongated cells the spindle is constrained close to the high stress axis.

Spatial effects of infiltrating T cells on neighbouring cancer cells and prognosis in stage III CRC patients.

Colorectal cancer (CRC) is one of the most frequently occurring cancers, but prognostic biomarkers identifying patients at risk of recurrence are still lacking. In this study, we aimed to investigate in more detail the spatial relationship between intratumoural T cells, cancer cells, and cancer cell hallmarks as prognostic biomarkers in stage III colorectal cancer patients. We conducted multiplexed imaging of 56 protein markers at single-cell resolution on resected fixed tissue from stage III CRC patients who received adjuvant 5-fluorouracil (5FU)-based chemotherapy. Images underwent segmentation for tumour, stroma, and immune cells, and cancer cell 'state' protein marker expression was quantified at a cellular level. We developed a Python package for estimation of spatial proximity, nearest neighbour analysis focusing on cancer cell-T-cell interactions at single-cell level. In our discovery cohort (Memorial Sloan Kettering samples), we processed 462 core samples (total number of cells: 1,669,228) from 221 adjuvant 5FU-treated stage III patients. The validation cohort (Huntsville Clearview Cancer Center samples) consisted of 272 samples (total number of cells: 853,398) from 98 stage III CRC patients. While there were trends for an association between the percentage of cytotoxic T cells (across the whole cancer core), it did not reach significance (discovery cohort: p = 0.07; validation cohort: p = 0.19). We next utilised our region-based nearest neighbour approach to determine the spatial relationships between cytotoxic T cells, helper T cells, and cancer cell clusters. In both cohorts, we found that shorter distance between cytotoxic T cells, T helper cells, and cancer cells was significantly associated with increased disease-free survival. An unsupervised trained model that clustered patients based on the median distance between immune cells and cancer cells, as well as protein expression profiles, successfully classified patients into low-risk and high-risk groups (discovery cohort: p = 0.01; validation cohort: p = 0.003). © 2024 The Author(s). The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.

Abnormal higher-order network interactions in Parkinson's disease visual hallucinations.

Visual hallucinations in Parkinson's disease can be viewed from a systems-level perspective, whereby dysfunctional communication between brain networks responsible for perception predisposes a person to hallucinate. To this end, abnormal functional interactions between higher-order and primary sensory networks have been implicated in the pathophysiology of visual hallucinations in Parkinson's disease, however the precise signatures remain to be determined. Dimensionality reduction techniques offer a novel means for simplifying the interpretation of multidimensional brain imaging data, identifying hierarchical patterns in the data that are driven by both within- and between-functional network changes. Here, we applied two complementary non-linear dimensionality reduction techniques-diffusion-map embedding and t-distributed stochastic neighbour embedding (t-SNE)-to resting state functional MRI data, in order to characterize the altered functional hierarchy associated with susceptibility to visual hallucinations. Our study involved 77 people with Parkinson's disease (31 with hallucinations; 46 without hallucinations) and 19 age-matched healthy control subjects. In patients with visual hallucinations, we found compression of the unimodal-heteromodal gradient consistent with increased functional integration between sensory and higher order networks. This was mirrored in a traditional functional connectivity analysis, which showed increased connectivity between the visual and default mode networks in the hallucinating group. Together, these results suggest a route by which higher-order regions may have excessive influence over earlier sensory processes, as proposed by theoretical models of hallucinations across disorders. By contrast, the t-SNE analysis identified distinct alterations in prefrontal regions, suggesting an additional layer of complexity in the functional brain network abnormalities implicated in hallucinations, which was not apparent in traditional functional connectivity analyses. Together, the results confirm abnormal brain organization associated with the hallucinating phenotype in Parkinson's disease and highlight the utility of applying convergent dimensionality reduction techniques to investigate complex clinical symptoms. In addition, the patterns we describe in Parkinson's disease converge with those seen in other conditions, suggesting that reduced hierarchical differentiation across sensory-perceptual systems may be a common transdiagnostic vulnerability in neuropsychiatric disorders with perceptual disturbances.

Nonlinear elasticity and short-range mechanical coupling govern the rate and symmetry of mouth opening in Hydra.

Hydra has a tubular bilayered epithelial body column with a dome-shaped head on one end and a foot on the other. Hydra lacks a permanent mouth: its head epithelium is sealed. Upon neuronal activation, a mouth opens at the apex of the head which can exceed the body column diameter in seconds, allowing Hydra to ingest prey larger than itself. While the kinematics of mouth opening are well characterized, the underlying mechanism is unknown. We show that Hydra mouth opening is generated by independent local contractions that require tissue-level coordination. We model the head epithelium as an active viscoelastic nonlinear spring network. The model reproduces the size, timescale and symmetry of mouth opening. It shows that radial contractions, travelling inwards from the outer boundary of the head, pull the mouth open. Nonlinear elasticity makes mouth opening larger and faster, contrary to expectations. The model correctly predicts changes in mouth shape in response to external forces. By generating innervated : nerve-free chimera in experiments and simulations, we show that nearest-neighbour mechanical signalling suffices to coordinate mouth opening. Hydra mouth opening shows that in the absence of long-range chemical or neuronal signals, short-range mechanical coupling is sufficient to produce long-range order in tissue deformations.

Integration of pairwise neighbor topologies and miRNA family and cluster attributes for miRNA-disease association prediction.

Identifying disease-related microRNAs (miRNAs) assists the understanding of disease pathogenesis. Existing research methods integrate multiple kinds of data related to miRNAs and diseases to infer candidate disease-related miRNAs. The attributes of miRNA nodes including their family and cluster belonging information, however, have not been deeply integrated. Besides, the learning of neighbor topology representation of a pair of miRNA and disease is a challenging issue. We present a disease-related miRNA prediction method by encoding and integrating multiple representations of miRNA and disease nodes learnt from the generative and adversarial perspective. We firstly construct a bilayer heterogeneous network of miRNA and disease nodes, and it contains multiple types of connections among these nodes, which reflect neighbor topology of miRNA-disease pairs, and the attributes of miRNA nodes, especially miRNA-related families and clusters. To learn enhanced pairwise neighbor topology, we propose a generative and adversarial model with a convolutional autoencoder-based generator to encode the low-dimensional topological representation of the miRNA-disease pair and multi-layer convolutional neural network-based discriminator to discriminate between the true and false neighbor topology embeddings. Besides, we design a novel feature category-level attention mechanism to learn the various importance of different features for final adaptive fusion and prediction. Comparison results with five miRNA-disease association methods demonstrated the superior performance of our model and technical contributions in terms of area under the receiver operating characteristic curve and area under the precision-recall curve. The results of recall rates confirmed that our model can find more actual miRNA-disease associations among top-ranked candidates. Case studies on three cancers further proved the ability to detect potential candidate miRNAs.

Development of a predictive machine learning model for pathogen profiles in patients with secondary immunodeficiency.

BACKGROUND: Secondary immunodeficiency can arise from various clinical conditions that include HIV infection, chronic diseases, malignancy and long-term use of immunosuppressives, which makes the suffering patients susceptible to all types of pathogenic infections. Other than HIV infection, the possible pathogen profiles in other aetiology-induced secondary immunodeficiency are largely unknown. METHODS: Medical records of the patients with secondary immunodeficiency caused by various aetiologies were collected from the First Affiliated Hospital of Nanchang University, China. Based on these records, models were developed with the machine learning method to predict the potential infectious pathogens that may inflict the patients with secondary immunodeficiency caused by various disease conditions other than HIV infection. RESULTS: Several metrics were used to evaluate the models' performance. A consistent conclusion can be drawn from all the metrics that Gradient Boosting Machine had the best performance with the highest accuracy at 91.01%, exceeding other models by 13.48, 7.14, and 4.49% respectively. CONCLUSIONS: The models developed in our study enable the prediction of potential infectious pathogens that may affect the patients with secondary immunodeficiency caused by various aetiologies except for HIV infection, which will help clinicians make a timely decision on antibiotic use before microorganism culture results return.

A major genetic locus in neighbours controls changes of gene expression and susceptibility in intraspecific rice mixtures.

Reports indicate that intraspecific neighbours alter the physiology of focal plants, and with a few exceptions, their molecular responses to neighbours are unknown. Recently, changes in susceptibility to pathogen resulting from such interactions were demonstrated, a phenomenon called neighbour-modulated susceptibility (NMS). However, the genetics of NMS and the associated molecular responses are largely unexplored. Here, we analysed in rice the modification of biomass and susceptibility to the blast fungus pathogen in the Kitaake focal genotype in the presence of 280 different neighbours. Using genome-wide association studies, we identified the loci in the neighbour that determine the response in Kitaake. Using a targeted transcriptomic approach, we characterized the molecular responses in focal plants co-cultivated with various neighbours inducing a reduction in susceptibility. Our study demonstrates that NMS is controlled by one major locus in the rice genome of its neighbour. Furthermore, we show that this locus can be associated with characteristic patterns of gene expression in focal plant. Finally, we propose an hypothesis where Pi could play a role in explaining this case of NMS. Our study sheds light on how plants affect the physiology in their neighbourhood and opens perspectives for understanding plant-plant interactions.

Pollinator foraging tactics have divergent consequences for the mating system of a tropical plant.

Resolving the consequences of pollinator foraging behaviour for plant mating systems is a fundamental challenge in evolutionary ecology. Pollinators may adopt particular foraging tactics: complete trapline foraging (repeated movements along a fixed route), sample-and-shift trapline foraging (a variable route that incorporates information from previous experiences) and territorial foraging (stochastic movements within a restricted area). Studies that integrate these pollinator foraging tactics with plant mating systems are generally lacking. We investigate the consequences of particular pollinator foraging tactics for Heliconia tortuosa. We combine parentage and sibship inference analysis with simulation modelling to: estimate mating system parameters; infer the foraging tactic adopted by the pollinators; and quantify the impact of pollinator foraging tactics on mating system parameters. We found high outcrossing rates, ubiquitous multiple paternity and a pronounced departure from near-neighbour mating. We also found that plants repeatedly receive pollen from a series of particular donors. We infer that the pollinators primarily adopt complete trapline foraging and occasionally engage in sample-and-shift trapline foraging. This enhances multiple paternity without a substantial increase in near-neighbour mating. The particular pollinator foraging tactics have divergent consequences for multiple paternity and near-neighbour mating. Thus, pollinator foraging behaviour is an important driver of the ecology and evolution of plant mating systems.

Measuring the randomness of micro- and nanostructure spatial distributions: Effects of scanning electron microscope image processing and analysis.

The quantitative characterisation of the degree of randomness and aggregation of surface micro- and nanostructures is critical to evaluate their effects on targeted functionalities. To this end, the methods of point pattern analysis (PPA), largely used in ecology and medical imaging, seem to provide a powerful toolset. However, the application of these techniques requires the extraction of the point pattern of nanostructures from their microscope images. In this work, we address the issue of the impact that Scanning Electron Microscope (SEM) image processing may have on the fundamental metric of PPA, that is, the Nearest Neighbour Index (NNI). Using typical SEM images of polymer micro- and nanostructures taken from secondary and backscattered electrons, we report the effects of the (a) noise filtering and (b) binarisation threshold on the value of NNI as well as the impact of the image finite size effects. Based on these results, we draw conclusions for the safe choice of SEM settings to provide accurate measurement of nanostructure randomness through NNI estimation.

KPCA-WRF-prediction of heart rate using deep feature fusion and machine learning classification with tuned weighted hyper-parameter.

The rapid advancement of technology such as stream processing technologies, deep-learning approaches, and artificial intelligence plays a prominent and vital role, to detect heart rate using a prediction model. However, the existing methods could not handle high -dimensional datasets, and deep feature learning to improvise the performance. Therefore, this work proposed a real-time heart rate prediction model, using K-nearest neighbour (KNN) adhered to the principle component analysis algorithm (PCA) and weighted random forest algorithm for feature fusion (KPCA-WRF) approach and deep CNN feature learning framework. The feature selection, from the fused features, was optimized by ant colony optimization (ACO) and particle swarm optimization (PSO) algorithm to enhance the selected fused features from deep CNN. The optimized features were reduced to low dimensions using the PCA algorithm. The significant straight heart rate features are plotted by capturing out nearest similar data point values using the algorithm. The fused features were then classified for aiding the training process. The weighted values are assigned to those tuned hyper parameters (feature matrix forms). The optimal path and continuity of the weighted feature representations are moved using the random forest algorithm, in K-fold validation iterations.

Nearest Neighbour Node Deployment Algorithm for Mobile Sensor Networks.

Many animal aggregations display remarkable collective coordinated movements on a large scale, which emerge as a result of distributed local decision-making by individuals. The recent advances in modelling the collective motion of animals through the utilisation of Nearest Neighbour rules, without the need for centralised coordination, resulted in the development of self-deployment algorithms in Mobile Sensor Networks (MSNs) to achieve various types of coverage essential for different applications. However, the energy consumption associated with sensor movement to achieve the desired coverage remains a significant concern for the majority of algorithms reported in the literature. In this paper, the Nearest Neighbour Node Deployment (NNND) algorithm is proposed to efficiently provide blanket coverage across a given area while minimising energy consumption and enhancing fault tolerance. In contrast to other algorithms that sequentially move sensors, NNND leverages the power of parallelism by employing multiple streams of sensor motions, each directed towards a distinct section of the area. The cohesion of each stream is maintained by adaptively choosing a leader for each stream while collision avoidance is also ensured. These properties contribute to minimising the travel distance within each stream, resulting in decreased energy consumption. Additionally, the utilisation of multiple leaders in NNND eliminates the presence of a single point of failure, hence enhancing the fault tolerance of the area coverage. The results of our extensive simulation study demonstrate that NNND not only achieves lower energy consumption but also a higher percentage of k-coverage.

Synthesis and validation algorithm followed by a weighting procedure to create a detailed anthropometric dataset for the German working-age population.

For the German working-age population no publicly available and detailed anthropometric raw dataset exists, although several studies have collected anthropometric datasets. Unfortunately, the publication of raw data may be restricted by data usage regulations. This study presents a synthesis and validation algorithm to create a virtual copy of an already existing dataset. A detailed anthropometric dataset from a regional epidemiological public-health study in Germany was used for the synthesis and validation algorithm. Results revealed only minor deviations within the validation process. Compared to the original dataset, the virtual dataset was statistically almost identical. In a next step, the virtual dataset was weighted to approximate nationally representative values. In summary, the computed unweighted and weighted virtual data can be published without restrictions and used for ergonomic designing. Furthermore, the synthesis and validation algorithm is suitable for the generation of virtual copies and can be applied to other detailed anthropometric datasets.

Data usage regulations may restrict the publication of anthropometric datasets. A synthesis and validation algorithm was developed which can be applied to existing anthropometric datasets to create a virtual copy that is almost identical and can be published. In the current study this algorithm was used for data from Germany.

Quantum and Quantum-Inspired Stereographic K Nearest-Neighbour Clustering.

Nearest-neighbour clustering is a simple yet powerful machine learning algorithm that finds natural application in the decoding of signals in classical optical-fibre communication systems. Quantum k-means clustering promises a speed-up over the classical k-means algorithm; however, it has been shown to not currently provide this speed-up for decoding optical-fibre signals due to the embedding of classical data, which introduces inaccuracies and slowdowns. Although still not achieving an exponential speed-up for NISQ implementations, this work proposes the generalised inverse stereographic projection as an improved embedding into the Bloch sphere for quantum distance estimation in k-nearest-neighbour clustering, which allows us to get closer to the classical performance. We also use the generalised inverse stereographic projection to develop an analogous classical clustering algorithm and benchmark its accuracy, runtime and convergence for decoding real-world experimental optical-fibre communication data. This proposed 'quantum-inspired' algorithm provides an improvement in both the accuracy and convergence rate with respect to the k-means algorithm. Hence, this work presents two main contributions. Firstly, we propose the general inverse stereographic projection into the Bloch sphere as a better embedding for quantum machine learning algorithms; here, we use the problem of clustering quadrature amplitude modulated optical-fibre signals as an example. Secondly, as a purely classical contribution inspired by the first contribution, we propose and benchmark the use of the general inverse stereographic projection and spherical centroid for clustering optical-fibre signals, showing that optimizing the radius yields a consistent improvement in accuracy and convergence rate.

NGS data vectorization, clustering, and finding key codons in SARS-CoV-2 variations.

The rapid global spread and dissemination of SARS-CoV-2 has provided the virus with numerous opportunities to develop several variants. Thus, it is critical to determine the degree of the variations and in which part of the virus those variations occurred. Therefore, in this study, methods that could be used to vectorize the sequence data, perform clustering analysis, and visualize the results were proposed using machine learning methods. To conduct this study, a total of 224,073 cases of SARS-CoV-2 sequence data were collected through NCBI and GISAID, and the data were visualized using dimensionality reduction and clustering analysis models such as T-SNE and DBSCAN. The SARS-CoV-2 virus, which was first detected, was distinguished from different variations, including Omicron and Delta, in the cluster results. Furthermore, it was possible to examine which codon changes in the spike protein caused the variants to be distinguished using feature importance extraction models such as Random Forest or Shapely Value. The proposed method has the advantage of being able to analyse and visualize a large amount of data at once compared to the existing tree-based sequence data analysis. The proposed method was able to identify and visualize significant changes between the SARS-CoV-2 virus, which was first detected in Wuhan, China, in December 2019, and the newly formed mutant virus group. As a result of clustering analysis using sequence data, it was possible to confirm the formation of clusters among various variants in a two-dimensional graph, and by extracting the importance of variables, it was possible to confirm which codon changes played a major role in distinguishing variants. Furthermore, since the proposed method can handle a variety of data sequences, it can be used for all kinds of diseases, including influenza and SARS-CoV-2. Therefore, the proposed method has the potential to become widely used for the effective analysis of disease variations.

Development of an image classification pipeline for atherosclerotic plaques assessment using supervised machine learning.

BACKGROUND: During atherosclerosis, the narrowing of the arterial lumen is observed through the accumulation of bio compounds and the formation of plaque within artery walls. A non-linear optical imaging modality (NLOM), coherent anti-stokes Raman scattering (CARS) microscopy, can be used to image lipid-rich structures commonly found in atherosclerotic plaques. By matching the lipid's molecular vibrational frequencies (CH bonds), it is possible to map the accumulation of lipid-rich structures without the need for exogenous labelling and/or processing of the samples. CARS allows for the visualization of the morphological features of plaque. In combination with supervised machine learning, CARS imaged morphological features can be used to characterize the progression of atherosclerotic plaques.  RESULTS: Based on a set of label-free CARS images of atherosclerotic plaques (i.e. foam cell clusters) from a Watanabe heritable hyperlipidemic rabbit model, we developed an automated pipeline to classify atherosclerotic lesions based on their major morphological features. Our method uses image preprocessing to first improve the quality of the CARS-imaged plaque, followed by the segmentation of the plaque using Otsu thresholding, marker-controlled watershed, K-means segmentation and a novel independent foam cell thresholding segmentation. To define relevant morphological features, 27 quantitative features were extracted and further refined by a novel coefficient of variation feature refinement method in accordance with filter-type feature selection. Refined morphological features were supplied into three supervised machine learning algorithms; K-nearest neighbour, support vector machine and decision tree classifier. The classification pipeline showcased the ability to exploit relevant plaque morphological features to accurately classify 3 pre-defined stages of atherosclerosis: early fatty streak development (EFS) and advancing atheroma (AA) with a greater than 85% class accuracy CONCLUSIONS: Through the combination of CARS microscopy and computational methods, a powerful classification tool was developed to identify the progression of atherosclerotic plaque in an automated manner. Using a curated dataset, the classification pipeline demonstrated the ability to differentiate between EFS, EF and AA. Thus, presenting the opportunity to classify the onset of atherosclerosis at an earlier stage of development.

Integrating eco-evolutionary dynamics and modern coexistence theory.

Community ecology typically assumes that competitive exclusion and species coexistence are unaffected by evolution on the time scale of ecological dynamics. However, recent studies suggest that rapid evolution operating concurrently with competition may enable species coexistence. Such findings necessitate general theory that incorporates the coexistence contributions of eco-evolutionary processes in parallel with purely ecological mechanisms and provides metrics for quantifying the role of evolution in shaping competitive outcomes in both modelling and empirical contexts. To foster the development of such theory, here we extend the interpretation of the two principal metrics of modern coexistence theory-niche and competitive ability differences-to systems where competitors evolve. We define eco-evolutionary versions of these metrics by considering how invading and resident species adapt to conspecific and heterospecific competitors. We show that the eco-evolutionary niche and competitive ability differences are sums of ecological and evolutionary processes, and that they accurately predict the potential for stable coexistence in previous theoretical studies of eco-evolutionary dynamics. Finally, we show how this theory frames recent empirical assessments of rapid evolution effects on species coexistence, and how empirical work and theory on species coexistence and eco-evolutionary dynamics can be further integrated.

Plant neighbours can make or break the disease transmission chain of a fungal root pathogen.

Biodiversity can reduce or increase disease transmission. These divergent effects suggest that community composition rather than diversity per se determines disease transmission. In natural plant communities, little is known about the functional roles of neighbouring plant species in belowground disease transmission. Here, we experimentally investigated disease transmission of a fungal root pathogen (Rhizoctonia solani) in two focal plant species in combinations with four neighbour species of two ages. We developed stochastic models to test the relative importance of two transmission-modifying mechanisms: (1) infected hosts serve as nutrient supply to increase hyphal growth, so that successful disease transmission is self-reinforcing; and (2) plant resistance increases during plant development. Neighbouring plants either reduced or increased disease transmission in the focal plants. These effects depended on neighbour age, but could not be explained by a simple dichotomy between hosts and nonhost neighbours. Model selection revealed that both transmission-modifying mechanisms are relevant and that focal host-neighbour interactions changed which mechanisms steered disease transmission rate. Our work shows that neighbour-induced shifts in the importance of these mechanisms across root networks either make or break disease transmission chains. Understanding how diversity affects disease transmission thus requires integrating interactions between focal and neighbour species and their pathogens.

The role of plant-soil feedback in long-term species coexistence cannot be predicted from its effects on plant performance.

BACKGROUND: Despite many studies on the importance of competition and plants' associations with mutualists and pathogens on plant performance and community organization, the joint effects of these two factors remain largely unexplored. Even less is known about how these joint effects vary through a plant's life in different environmental conditions and how they contribute to the long-term coexistence of species. METHODS: We investigated the role of plant-soil feedback (PSF) in intra- and interspecific competition, using two co-occurring dry grassland species as models. A two-phase PSF experiment was used. In the first phase, soil was conditioned by the two plant species. In the second, we assessed the effect of soil conditioning, competition and drought stress on seedling establishment, plant growth in the first and second vegetation season, and fruit production. We also estimated effects of different treatments on overall population growth rates and predicted the species' potential coexistence. RESULTS: Soil conditioning played a more important role in the early stages of the plants' life (seedling establishment and early growth) than competition. Specifically, we found strong negative intraspecific PSF for biomass production in the first year in both species. Although the effects of soil conditioning persisted in later stages of plant's life, competition and drought stress became more important. Surprisingly, models predicting species coexistence contrasted with the effects on individual life stages, showing that our model species benefit from their self-conditioned soil in the long run. CONCLUSIONS: We provide evidence that the effects of PSF vary through plants' life stages. Our study suggests that we cannot easily predict the effects of soil conditioning on long-term coexistence of species using data only on performance at a single time as commonly done in PSF studies. We also show the importance of using as realistic environmental conditions as possible (such as drought stress experienced in dry grasslands) to draw reasonable conclusions on species coexistence.