Learning agile soccer skills for a bipedal robot with deep reinforcement learning.

We investigated whether deep reinforcement learning (deep RL) is able to synthesize sophisticated and safe movement skills for a low-cost, miniature humanoid robot that can be composed into complex behavioral strategies. We used deep RL to train a humanoid robot to play a simplified one-versus-one soccer game. The resulting agent exhibits robust and dynamic movement skills, such as rapid fall recovery, walking, turning, and kicking, and it transitions between them in a smooth and efficient manner. It also learned to anticipate ball movements and block opponent shots. The agent's tactical behavior adapts to specific game contexts in a way that would be impractical to manually design. Our agent was trained in simulation and transferred to real robots zero-shot. A combination of sufficiently high-frequency control, targeted dynamics randomization, and perturbations during training enabled good-quality transfer. In experiments, the agent walked 181% faster, turned 302% faster, took 63% less time to get up, and kicked a ball 34% faster than a scripted baseline.

Crop growth dynamics: Fast automatic analysis of LiDAR images in field-plot experiments by specialized software ALFA.

Repeated measurements of crop height to observe plant growth dynamics in real field conditions represent a challenging task. Although there are ways to collect data using sensors on UAV systems, proper data processing and analysis are the key to reliable results. As there is need for specialized software solutions for agricultural research and breeding purposes, we present here a fast algorithm ALFA for the processing of UAV LiDAR derived point-clouds to extract the information on crop height at many individual cereal field-plots at multiple time points. Seven scanning flights were performed over 3 blocks of experimental barley field plots between April and June 2021. Resulting point-clouds were processed by the new algorithm ALFA. The software converts point-cloud data into a digital image and extracts the traits of interest-the median crop height at individual field plots. The entire analysis of 144 field plots of dimension 80 x 33 meters measured at 7 time points (approx. 100 million LiDAR points) takes about 3 minutes at a standard PC. The Root Mean Square Deviation of the software-computed crop height from the manual measurement is 5.7 cm. Logistic growth model is fitted to the measured data by means of nonlinear regression. Three different ways of crop-height data visualization are provided by the software to enable further analysis of the variability in growth parameters. We show that the presented software solution is a fast and reliable tool for automatic extraction of plant height from LiDAR images of individual field-plots. We offer this tool freely to the scientific community for non-commercial use.

Inferring couplings in networks across order-disorder phase transitions.

Statistical inference is central to many scientific endeavors, yet how it works remains unresolved. Answering this requires a quantitative understanding of the intrinsic interplay between statistical models, inference methods, and the structure in the data. To this end, we characterize the efficacy of direct coupling analysis (DCA) - a highly successful method for analyzing amino acid sequence data-in inferring pairwise interactions from samples of ferromagnetic Ising models on random graphs. Our approach allows for physically motivated exploration of qualitatively distinct data regimes separated by phase transitions. We show that inference quality depends strongly on the nature of data-generating distributions: optimal accuracy occurs at an intermediate temperature where the detrimental effects from macroscopic order and thermal noise are minimal. Importantly our results indicate that DCA does not always outperform its local-statistics-based predecessors; while DCA excels at low temperatures, it becomes inferior to simple correlation thresholding at virtually all temperatures when data are limited. Our findings offer insights into the regime in which DCA operates so successfully, and more broadly, how inference interacts with the structure in the data.

Bayesian approach for analysis of time-to-event data in plant biology.

BACKGROUND: Plants, like all living organisms, metamorphose their bodies during their lifetime. All the developmental and growth events in a plant's life are connected to specific points in time, be it seed germination, seedling emergence, the appearance of the first leaf, heading, flowering, fruit ripening, wilting, or death. The onset of automated phenotyping methods has brought an explosion of such time-to-event data. Unfortunately, it has not been matched by an explosion of adequate data analysis methods. RESULTS AND DISCUSSION: In this paper, we introduce the Bayesian approach towards time-to-event data in plant biology. As a model example, we use seedling emergence data of maize under control and stress conditions but the Bayesian approach is suitable for any time-to-event data (see the examples above). In the proposed framework, we are able to answer key questions regarding plant emergence such as these: (1) Do seedlings treated with compound A emerge earlier than the control seedlings? (2) What is the probability of compound A increasing seedling emergence by at least 5 percent? CONCLUSION: Proper data analysis is a fundamental task of general interest in life sciences. Here, we present a novel method for the analysis of time-to-event data which is applicable to many plant developmental parameters measured in field or in laboratory conditions. In contrast to recent and classical approaches, our Bayesian computational method properly handles uncertainty in time-to-event data and it is capable to reliably answer questions that are difficult to address by classical methods.

A Novel Image-Based Screening Method to Study Water-Deficit Response and Recovery of Barley Populations Using Canopy Dynamics Phenotyping and Simple Metabolite Profiling.

Plant phenotyping platforms offer automated, fast scoring of traits that simplify the selection of varieties that are more competitive under stress conditions. However, indoor phenotyping methods are frequently based on the analysis of plant growth in individual pots. We present a reproducible indoor phenotyping method for screening young barley populations under water stress conditions and after subsequent rewatering. The method is based on a simple read-out of data using RGB imaging, projected canopy height, as a useful feature for indirectly following the kinetics of growth and water loss in a population of barley. A total of 47 variables including 15 traits and 32 biochemical metabolites measured (morphometric parameters, chlorophyll fluorescence imaging, quantification of stress-related metabolites; amino acids and polyamines, and enzymatic activities) were used to validate the method. The study allowed the identification of metabolites related to water stress response and recovery. Specifically, we found that cadaverine (Cad), 1,3-aminopropane (DAP), tryptamine (Tryp), and tyramine (Tyra) were the major contributors to the water stress response, whereas Cad, DAP, and Tyra, but not Tryp, remained at higher levels in the stressed plants even after rewatering. In this work, we designed, optimized and validated a non-invasive image-based method for automated screening of potential water stress tolerance genotypes in barley populations. We demonstrated the applicability of the method using transgenic barley lines with different sensitivity to drought stress showing that combining canopy height and the metabolite profile we can discriminate tolerant from sensitive genotypes. We showed that the projected canopy height a sensitive trait that truly reflects other invasively studied morphological, physiological, and metabolic traits and that our presented methodological setup can be easily applicable for large-scale screenings in low-cost systems equipped with a simple RGB camera. We believe that our approach will contribute to accelerate the study and understanding of the plant water stress response and recovery capacity in crops, such as barley.

Plant responses to fungal volatiles involve global posttranslational thiol redox proteome changes that affect photosynthesis.

Microorganisms produce volatile compounds (VCs) that promote plant growth and photosynthesis through complex mechanisms involving cytokinin (CK) and abscisic acid (ABA). We hypothesized that plants' responses to microbial VCs involve posttranslational modifications of the thiol redox proteome through action of plastidial NADPH-dependent thioredoxin reductase C (NTRC), which regulates chloroplast redox status via its functional relationship with 2-Cys peroxiredoxins. To test this hypothesis, we analysed developmental, metabolic, hormonal, genetic, and redox proteomic responses of wild-type (WT) plants and a NTRC knockout mutant (ntrc) to VCs emitted by the phytopathogen Alternaria alternata. Fungal VC-promoted growth, changes in root architecture, shifts in expression of VC-responsive CK- and ABA-regulated genes, and increases in photosynthetic capacity were substantially weaker in ntrc plants than in WT plants. As in WT plants, fungal VCs strongly promoted growth, chlorophyll accumulation, and photosynthesis in ntrc-Δ2cp plants with reduced 2-Cys peroxiredoxin expression. OxiTRAQ-based quantitative and site-specific redox proteomic analyses revealed that VCs promote global reduction of the thiol redox proteome (especially of photosynthesis-related proteins) of WT leaves but its oxidation in ntrc leaves. Our findings show that NTRC is an important mediator of plant responses to microbial VCs through mechanisms involving global thiol redox proteome changes that affect photosynthesis.

Characterization of Biostimulant Mode of Action Using Novel Multi-Trait High-Throughput Screening of Arabidopsis Germination and Rosette Growth.

Environmental stresses have a significant effect on agricultural crop productivity worldwide. Exposure of seeds to abiotic stresses, such as salinity among others, results in lower seed viability, reduced germination, and poor seedling establishment. Alternative agronomic practices, e.g., the use of plant biostimulants, have attracted considerable interest from the scientific community and commercial enterprises. Biostimulants, i.e., products of biological origin (including bacteria, fungi, seaweeds, higher plants, or animals) have significant potential for (i) improving physiological processes in plants and (ii) stimulating germination, growth and stress tolerance. However, biostimulants are diverse, and can range from single compounds to complex matrices with different groups of bioactive components that have only been partly characterized. Due to the complex mixtures of biologically active compounds present in biostimulants, efficient methods for characterizing their potential mode of action are needed. In this study, we report the development of a novel complex approach to biological activity testing, based on multi-trait high-throughput screening (MTHTS) of Arabidopsis characteristics. These include the in vitro germination rate, early seedling establishment capacity, growth capacity under stress and stress response. The method is suitable for identifying new biostimulants and characterizing their mode of action. Representatives of compatible solutes such as amino acids and polyamines known to be present in many of the biostimulant irrespective of their origin, i.e., well-established biostimulants that enhance stress tolerance and crop productivity, were used for the assay optimization and validation. The selected compounds were applied through seed priming over a broad concentration range and the effect was investigated simultaneously under control, moderate stress and severe salt stress conditions. The new MTHTS approach represents a powerful tool in the field of biostimulant research and development and offers direct classification of the biostimulants mode of action into three categories: (1) plant growth promotors/inhibitors, (2) stress alleviators, and (3) combined action.

Cytokinins: Their Impact on Molecular and Growth Responses to Drought Stress and Recovery in Arabidopsis.

Our phenotyping and hormonal study has characterized the role of cytokinins (CK) in the drought and recovery responses of Arabidopsis thaliana. CK down-regulation was achieved by overexpression of the gene for CK deactivating enzyme cytokinin oxidase/dehydrogenase (CKX): constitutive (35S:CKX) or at the stress onset using a dexamethasone-inducible pOp/LhGR promoter (DEX:CKX). The 35S:CKX plants exhibited slow ontogenesis and higher expression levels of stress-associated genes, e.g., AtP5CS1, already at well-watered conditions. CK down-regulation resulted during drought in higher stress tolerance (indicated by relatively low up-regulation of the expression of drought stress marker gene AtRD29B) accompanied with lower leaf water loss. Nevertheless, these plants exhibited slow and delayed recovery after re-watering. CK levels were increased at the stress onset by stimulation of the expression of CK biosynthetic gene isopentenyl transferase (ipt) (DEX:IPT) or by application of exogenous CK meta-topolin. After water withdrawal, long-term CK elevation resulted in higher water loss in comparison with CKX transformants as well as with plants overexpressing ipt driven by senescence-inducible SAG12 promoter (SAG:IPT), which gradually enhanced CKs during the stress progression. In all cases, CK up-regulation resulted in fast and more vigorous recovery. All drought-stressed plants exhibited growth suppression associated with elevation of abscisic acid and decrease of auxins and active CKs (with the exception of SAG:IPT plants). Apart from the ipt overexpressers, also increase of jasmonic and salicylic acid was found.

An Automated Method for High-Throughput Screening of Arabidopsis Rosette Growth in Multi-Well Plates and Its Validation in Stress Conditions.

High-throughput plant phenotyping platforms provide new possibilities for automated, fast scoring of several plant growth and development traits, followed over time using non-invasive sensors. Using Arabidopsis as a model offers important advantages for high-throughput screening with the opportunity to extrapolate the results obtained to other crops of commercial interest. In this study we describe the development of a highly reproducible high-throughput Arabidopsis in vitro bioassay established using our OloPhen platform, suitable for analysis of rosette growth in multi-well plates. This method was successfully validated on example of multivariate analysis of Arabidopsis rosette growth in different salt concentrations and the interaction with varying nutritional composition of the growth medium. Several traits such as changes in the rosette area, relative growth rate, survival rate and homogeneity of the population are scored using fully automated RGB imaging and subsequent image analysis. The assay can be used for fast screening of the biological activity of chemical libraries, phenotypes of transgenic or recombinant inbred lines, or to search for potential quantitative trait loci. It is especially valuable for selecting genotypes or growth conditions that improve plant stress tolerance.

Probabilistic models for neural populations that naturally capture global coupling and criticality.

Advances in multi-unit recordings pave the way for statistical modeling of activity patterns in large neural populations. Recent studies have shown that the summed activity of all neurons strongly shapes the population response. A separate recent finding has been that neural populations also exhibit criticality, an anomalously large dynamic range for the probabilities of different population activity patterns. Motivated by these two observations, we introduce a class of probabilistic models which takes into account the prior knowledge that the neural population could be globally coupled and close to critical. These models consist of an energy function which parametrizes interactions between small groups of neurons, and an arbitrary positive, strictly increasing, and twice differentiable function which maps the energy of a population pattern to its probability. We show that: 1) augmenting a pairwise Ising model with a nonlinearity yields an accurate description of the activity of retinal ganglion cells which outperforms previous models based on the summed activity of neurons; 2) prior knowledge that the population is critical translates to prior expectations about the shape of the nonlinearity; 3) the nonlinearity admits an interpretation in terms of a continuous latent variable globally coupling the system whose distribution we can infer from data. Our method is independent of the underlying system's state space; hence, it can be applied to other systems such as natural scenes or amino acid sequences of proteins which are also known to exhibit criticality.

To Stimulate or Inhibit? That Is the Question for the Function of Abscisic Acid.

Physiologically, abscisic acid (ABA) is believed to be a general inhibitor of plant growth, including during the crucial early development of seedlings. However, this view contradicts many reports of stimulatory effects of ABA that, so far, have not been considered in the debate concerning ABA's function in plant development. To address this apparent contradiction, we propose a hypothetical mechanism to explain how ABA might contribute to the promotion of cell expansion. We wish to overturn conventional views on ABA's role during juvenile plant development and put forward the idea that, as for other phytohormones, the role of ABA is determined by dose and sensitivity and ranges from stimulatory to inhibitory effects.

Arabidopsis Responds to Alternaria alternata Volatiles by Triggering Plastid Phosphoglucose Isomerase-Independent Mechanisms.

Volatile compounds (VCs) emitted by phylogenetically diverse microorganisms (including plant pathogens and microbes that do not normally interact mutualistically with plants) promote photosynthesis, growth, and the accumulation of high levels of starch in leaves through cytokinin (CK)-regulated processes. In Arabidopsis (Arabidopsis thaliana) plants not exposed to VCs, plastidic phosphoglucose isomerase (pPGI) acts as an important determinant of photosynthesis and growth, likely as a consequence of its involvement in the synthesis of plastidic CKs in roots. Moreover, this enzyme plays an important role in connecting the Calvin-Benson cycle with the starch biosynthetic pathway in leaves. To elucidate the mechanisms involved in the responses of plants to microbial VCs and to investigate the extent of pPGI involvement, we characterized pPGI-null pgi1-2 Arabidopsis plants cultured in the presence or absence of VCs emitted by Alternaria alternata We found that volatile emissions from this fungal phytopathogen promote growth, photosynthesis, and the accumulation of plastidic CKs in pgi1-2 leaves. Notably, the mesophyll cells of pgi1-2 leaves accumulated exceptionally high levels of starch following VC exposure. Proteomic analyses revealed that VCs promote global changes in the expression of proteins involved in photosynthesis, starch metabolism, and growth that can account for the observed responses in pgi1-2 plants. The overall data show that Arabidopsis plants can respond to VCs emitted by phytopathogenic microorganisms by triggering pPGI-independent mechanisms.

Volatile compounds emitted by diverse phytopathogenic microorganisms promote plant growth and flowering through cytokinin action.

It is known that volatile emissions from some beneficial rhizosphere microorganisms promote plant growth. Here we show that volatile compounds (VCs) emitted by phylogenetically diverse rhizosphere and non-rhizhosphere bacteria and fungi (including plant pathogens and microbes that do not normally interact mutualistically with plants) promote growth and flowering of various plant species, including crops. In Arabidopsis plants exposed to VCs emitted by the phytopathogen Alternaria alternata, changes included enhancement of photosynthesis and accumulation of high levels of cytokinins (CKs) and sugars. Evidence obtained using transgenic Arabidopsis plants with altered CK status show that CKs play essential roles in this phenomenon, because growth and flowering responses to the VCs were reduced in mutants with CK-deficiency (35S:AtCKX1) or low receptor sensitivity (ahk2/3). Further, we demonstrate that the plant responses to fungal VCs are light-dependent. Transcriptomic analyses of Arabidopsis leaves exposed to A. alternata VCs revealed changes in the expression of light- and CK-responsive genes involved in photosynthesis, growth and flowering. Notably, many genes differentially expressed in plants treated with fungal VCs were also differentially expressed in plants exposed to VCs emitted by the plant growth promoting rhizobacterium Bacillus subtilis GB03, suggesting that plants react to microbial VCs through highly conserved regulatory mechanisms.

Spatio-temporal changes in endogenous abscisic acid contents during etiolated growth and photomorphogenesis in tomato seedlings.

The role of abscisic acid (ABA) during early development was investigated in tomato seedlings. The endogenous content of ABA in particular organs was analyzed in seedlings grown in the dark and under blue light. Our results showed that in dark-grown seedlings, the ABA accumulation was maximal in the cotyledons and elongation zone of hypocotyl, whereas under blue-light, the ABA content was distinctly reduced. Our data are consistent with the conclusion that ABA promotes the growth of etiolated seedlings and the results suggest that ABA plays an inhibitory role in de-etiolation and photomorphogenesis in tomato.

Automated phenotyping of plant shoots using imaging methods for analysis of plant stress responses - a review.

Current methods of in-house plant phenotyping are providing a powerful new tool for plant biology studies. The self-constructed and commercial platforms established in the last few years, employ non-destructive methods and measurements on a large and high-throughput scale. The platforms offer to certain extent, automated measurements, using either simple single sensor analysis, or advanced integrative simultaneous analysis by multiple sensors. However, due to the complexity of the approaches used, it is not always clear what such forms of plant phenotyping can offer the potential end-user, i.e. plant biologist. This review focuses on imaging methods used in the phenotyping of plant shoots including a brief survey of the sensors used. To open up this topic to a broader audience, we provide here a simple introduction to the principles of automated non-destructive analysis, namely RGB, chlorophyll fluorescence, thermal and hyperspectral imaging. We further on present an overview on how and to which extent, the automated integrative in-house phenotyping platforms have been used recently to study the responses of plants to various changing environments.

Plastidic phosphoglucose isomerase is an important determinant of starch accumulation in mesophyll cells, growth, photosynthetic capacity, and biosynthesis of plastidic cytokinins in Arabidopsis.

Phosphoglucose isomerase (PGI) catalyzes the reversible isomerization of glucose-6-phosphate and fructose-6-phosphate. It is involved in glycolysis and in the regeneration of glucose-6-P molecules in the oxidative pentose phosphate pathway (OPPP). In chloroplasts of illuminated mesophyll cells PGI also connects the Calvin-Benson cycle with the starch biosynthetic pathway. In this work we isolated pgi1-3, a mutant totally lacking pPGI activity as a consequence of aberrant intron splicing of the pPGI encoding gene, PGI1. Starch content in pgi1-3 source leaves was ca. 10-15% of that of wild type (WT) leaves, which was similar to that of leaves of pgi1-2, a T-DNA insertion pPGI null mutant. Starch deficiency of pgi1 leaves could be reverted by the introduction of a sex1 null mutation impeding ß-amylolytic starch breakdown. Although previous studies showed that starch granules of pgi1-2 leaves are restricted to both bundle sheath cells adjacent to the mesophyll and stomata guard cells, microscopy analyses carried out in this work revealed the presence of starch granules in the chloroplasts of pgi1-2 and pgi1-3 mesophyll cells. RT-PCR analyses showed high expression levels of plastidic and extra-plastidic ß-amylase encoding genes in pgi1 leaves, which was accompanied by increased ß-amylase activity. Both pgi1-2 and pgi1-3 mutants displayed slow growth and reduced photosynthetic capacity phenotypes even under continuous light conditions. Metabolic analyses revealed that the adenylate energy charge and the NAD(P)H/NAD(P) ratios in pgi1 leaves were lower than those of WT leaves. These analyses also revealed that the content of plastidic 2-C-methyl-D-erythritol 4-phosphate (MEP)-pathway derived cytokinins (CKs) in pgi1 leaves were exceedingly lower than in WT leaves. Noteworthy, exogenous application of CKs largely reverted the low starch content phenotype of pgi1 leaves. The overall data show that pPGI is an important determinant of photosynthesis, energy status, growth and starch accumulation in mesophyll cells likely as a consequence of its involvement in the production of OPPP/glycolysis intermediates necessary for the synthesis of plastidic MEP-pathway derived hormones such as CKs.

Automated integrative high-throughput phenotyping of plant shoots: a case study of the cold-tolerance of pea (Pisum sativum L.).

BACKGROUND: Recently emerging approaches to high-throughput plant phenotyping have discovered their importance as tools in unravelling the complex questions of plant growth, development and response to the environment, both in basic and applied science. High-throughput methods have been also used to study plant responses to various types of biotic and abiotic stresses (drought, heat, salinity, nutrient-starving, UV light) but only rarely to cold tolerance. RESULTS: We present here an experimental procedure of integrative high-throughput in-house phenotyping of plant shoots employing automated simultaneous analyses of shoot biomass and photosystem II efficiency to study the cold tolerance of pea (Pisum sativum L.). For this purpose, we developed new software for automatic RGB image analysis, evaluated various parameters of chlorophyll fluorescence obtained from kinetic chlorophyll fluorescence imaging, and performed an experiment in which the growth and photosynthetic activity of two different pea cultivars were followed during cold acclimation. The data obtained from the automated RGB imaging were validated through correlation of pixel based shoot area with measurement of the shoot fresh weight. Further, data obtained from automated chlorophyll fluorescence imaging analysis were compared with chlorophyll fluorescence parameters measured by a non-imaging chlorophyll fluorometer. In both cases, high correlation was obtained, confirming the reliability of the procedure described. CONCLUSIONS: This study of the response of two pea cultivars to cold stress confirmed that our procedure may have important application, not only for selection of cold-sensitive/tolerant varieties of pea, but also for studies of plant cold-response strategies in general. The approach, provides a very broad tool for the morphological and physiological selection of parameters which correspond to shoot growth and the efficiency of photosystem II, and is thus applicable in studies of various plant species and crops.

Endogenous abscisic acid promotes hypocotyl growth and affects endoreduplication during dark-induced growth in tomato (Solanum lycopersicum L.).

Dark-induced growth (skotomorphogenesis) is primarily characterized by rapid elongation of the hypocotyl. We have studied the role of abscisic acid (ABA) during the development of young tomato (Solanum lycopersicum L.) seedlings. We observed that ABA deficiency caused a reduction in hypocotyl growth at the level of cell elongation and that the growth in ABA-deficient plants could be improved by treatment with exogenous ABA, through which the plants show a concentration dependent response. In addition, ABA accumulated in dark-grown tomato seedlings that grew rapidly, whereas seedlings grown under blue light exhibited low growth rates and accumulated less ABA. We demonstrated that ABA promotes DNA endoreduplication by enhancing the expression of the genes encoding inhibitors of cyclin-dependent kinases SlKRP1 and SlKRP3 and by reducing cytokinin levels. These data were supported by the expression analysis of the genes which encode enzymes involved in ABA and CK metabolism. Our results show that ABA is essential for the process of hypocotyl elongation and that appropriate control of the endogenous level of ABA is required in order to drive the growth of etiolated seedlings.

Evolutionary dynamics of infectious diseases in finite populations.

In infectious disease epidemiology the basic reproductive ratio, R0, is defined as the average number of new infections caused by a single infected individual in a fully susceptible population. Many models describing competition for hosts between non-interacting pathogen strains in an infinite population lead to the conclusion that selection favors invasion of new strains if and only if they have higher R0 values than the resident. Here we demonstrate that this picture fails in finite populations. Using a simple stochastic SIS model, we show that in general there is no analogous optimization principle. We find that successive invasions may in some cases lead to strains that infect a smaller fraction of the host population, and that mutually invasible pathogen strains exist. In the limit of weak selection we demonstrate that an optimization principle does exist, although it differs from R0 maximization. For strains with very large R0, we derive an expression for this local fitness function and use it to establish a lower bound for the error caused by neglecting stochastic effects. Furthermore, we apply this weak selection limit to investigate the selection dynamics in the presence of a trade-off between the virulence and the transmission rate of a pathogen.