Mathematical discoveries from program search with large language models.

Large language models (LLMs) have demonstrated tremendous capabilities in solving complex tasks, from quantitative reasoning to understanding natural language. However, LLMs sometimes suffer from confabulations (or hallucinations), which can result in them making plausible but incorrect statements1,2. This hinders the use of current large models in scientific discovery. Here we introduce FunSearch (short for searching in the function space), an evolutionary procedure based on pairing a pretrained LLM with a systematic evaluator. We demonstrate the effectiveness of this approach to surpass the best-known results in important problems, pushing the boundary of existing LLM-based approaches3. Applying FunSearch to a central problem in extremal combinatorics-the cap set problem-we discover new constructions of large cap sets going beyond the best-known ones, both in finite dimensional and asymptotic cases. This shows that it is possible to make discoveries for established open problems using LLMs. We showcase the generality of FunSearch by applying it to an algorithmic problem, online bin packing, finding new heuristics that improve on widely used baselines. In contrast to most computer search approaches, FunSearch searches for programs that describe how to solve a problem, rather than what the solution is. Beyond being an effective and scalable strategy, discovered programs tend to be more interpretable than raw solutions, enabling feedback loops between domain experts and FunSearch, and the deployment of such programs in real-world applications.

Discovering faster matrix multiplication algorithms with reinforcement learning.

Improving the efficiency of algorithms for fundamental computations can have a widespread impact, as it can affect the overall speed of a large amount of computations. Matrix multiplication is one such primitive task, occurring in many systems-from neural networks to scientific computing routines. The automatic discovery of algorithms using machine learning offers the prospect of reaching beyond human intuition and outperforming the current best human-designed algorithms. However, automating the algorithm discovery procedure is intricate, as the space of possible algorithms is enormous. Here we report a deep reinforcement learning approach based on AlphaZero1 for discovering efficient and provably correct algorithms for the multiplication of arbitrary matrices. Our agent, AlphaTensor, is trained to play a single-player game where the objective is finding tensor decompositions within a finite factor space. AlphaTensor discovered algorithms that outperform the state-of-the-art complexity for many matrix sizes. Particularly relevant is the case of 4 × 4 matrices in a finite field, where AlphaTensor's algorithm improves on Strassen's two-level algorithm for the first time, to our knowledge, since its discovery 50 years ago2. We further showcase the flexibility of AlphaTensor through different use-cases: algorithms with state-of-the-art complexity for structured matrix multiplication and improved practical efficiency by optimizing matrix multiplication for runtime on specific hardware. Our results highlight AlphaTensor's ability to accelerate the process of algorithmic discovery on a range of problems, and to optimize for different criteria.

Spatial Dependency in Stubble-Borne Pyrenophora teres f. teres and Influence of Sample Support Size on DNA Concentration and Fungicide Resistance Frequency.

Fungicide resistance in foliar fungal pathogens is an increasing challenge to crop production. Yield impacts due to loss of fungicide efficacy may be reduced through effective surveillance and appropriate management intervention. For stubble-borne pathogens, off-season crop residues may be used to monitor fungicide resistance to inform pre-planting decisions; however, appropriate sampling strategies and support sizes for crop residues have not previously been considered. Here, we used Pyrenophora teres f. teres (Ptt) with resistance to demethylase inhibitor fungicides as a model system to assess spatial dependency and to compare the effects of different sampling strategies and support sizes on pathogen density (Ptt DNA concentration) and the frequency of fungicide resistance mutation. The results showed that sampling strategies (hand-picked versus raked) did not affect estimates of pathogen density or fungicide resistance frequency; however, sample variances were lower from raked samples. The effects of differing sample support size, as the size of the collection area (1.2, 8.6, or 60 m2), on fungicide resistance frequency were not evident (P > 0.05). However, measures of pathogen density increased with area size (P < 0.05); the 60 m2 area yielded the highest Ptt DNA concentration and produced the lowest number of pathogen-absent samples. Sample variances for pathogen density and fungicide resistance frequency were generally homogeneous between area sizes. The pattern of pathogen density was spatially independent; however, spatial dependency was identified for fungicide resistance frequency, with a range of 110 m, in one of the two fields surveyed. Collectively, the results inform designs for monitoring of fungicide resistance in stubble-borne pathogens.

Optimized Sample Processing Pipeline for PCR-Based Fungicide Resistance Quantification of Stubble-Borne Fungal Pathogens.

Globally, yield losses associated with failed crop protection due to fungicide-resistant pathogens present an increasing problem. For stubble-borne pathogens, assessment of crop residues during the off-season could provide early fungicide resistance quantification for informed management decisions to mitigate yield losses. However, stubble assessment is hampered by assay inhibitors that are derived from decaying organic matter. To overcome assay inhibition from weathered stubble samples, we used a systems approach to quantify the frequency of resistance to demethylase inhibitor fungicides of the barley pathogen Pyrenophora teres f. teres. The system canvassed (i) 10 ball-milling conditions; (ii) four DNA extraction methodologies; and (iii) three column purification techniques for the provision of sufficient yield, quality, and purity of fungal DNA for a PCR-based fungicide resistance assay. Results show that DNA quantity and purity differed within each of the above three categories, with the optimized pipeline being (i) ball-milling samples in a 50-ml stainless steel canister for 5 min using a 20-mm ball at 30 revolutions s-1; (ii) a modified Brandfass method (extracted 64% more DNA than other methods assessed); and (iii) use of silica resin columns for the highest DNA concentration with optimal DNA purity. The chip-digital PCR assay, which quantified fungicide resistance from field samples, was unaffected by the DNA extraction method or purification technique, provided that thresholds of template quantity and purity were satisfied. In summary, this study has developed molecular pipeline options for pathogen fungicide resistance quantification from cereal stubbles, which can guide management for improved crop protection outcomes.

Predictors of changing patterns of adherence to containment measures during the early stage of COVID-19 pandemic: an international longitudinal study.

BACKGROUND: Identifying common factors that affect public adherence to COVID-19 containment measures can directly inform the development of official public health communication strategies. The present international longitudinal study aimed to examine whether prosociality, together with other theoretically derived motivating factors (self-efficacy, perceived susceptibility and severity of COVID-19, perceived social support) predict the change in adherence to COVID-19 containment strategies. METHOD: In wave 1 of data collection, adults from eight geographical regions completed online surveys beginning in April 2020, and wave 2 began in June and ended in September 2020. Hypothesized predictors included prosociality, self-efficacy in following COVID-19 containment measures, perceived susceptibility to COVID-19, perceived severity of COVID-19 and perceived social support. Baseline covariates included age, sex, history of COVID-19 infection and geographical regions. Participants who reported adhering to specific containment measures, including physical distancing, avoidance of non-essential travel and hand hygiene, were classified as adherence. The dependent variable was the category of adherence, which was constructed based on changes in adherence across the survey period and included four categories: non-adherence, less adherence, greater adherence and sustained adherence (which was designated as the reference category). RESULTS: In total, 2189 adult participants (82% female, 57.2% aged 31-59 years) from East Asia (217 [9.7%]), West Asia (246 [11.2%]), North and South America (131 [6.0%]), Northern Europe (600 [27.4%]), Western Europe (322 [14.7%]), Southern Europe (433 [19.8%]), Eastern Europe (148 [6.8%]) and other regions (96 [4.4%]) were analyzed. Adjusted multinomial logistic regression analyses showed that prosociality, self-efficacy, perceived susceptibility and severity of COVID-19 were significant factors affecting adherence. Participants with greater self-efficacy at wave 1 were less likely to become non-adherence at wave 2 by 26% (adjusted odds ratio [aOR], 0.74; 95% CI, 0.71 to 0.77; P < .001), while those with greater prosociality at wave 1 were less likely to become less adherence at wave 2 by 23% (aOR, 0.77; 95% CI, 0.75 to 0.79; P = .04). CONCLUSIONS: This study provides evidence that in addition to emphasizing the potential severity of COVID-19 and the potential susceptibility to contact with the virus, fostering self-efficacy in following containment strategies and prosociality appears to be a viable public health education or communication strategy to combat COVID-19.

Psychometric properties of the spanish version of the Valuing Questionnaire in Colombian clinical and nonclinical samples.

OBJECTIVE: To examine the psychometric properties of the Spanish version of the Valuing Questionnaire (VQ) in Colombian clinical and nonclinical samples. METHOD: The VQ was administered to a total sample of 1820 participants, which included undergraduates (N = 762), general population (N = 724), and a clinical sample (N = 334). The questionnaire packages included measures of experiential avoidance, cognitive fusion, mindfulness, life satisfaction, and psychological difficulties. RESULTS: Across the different samples, internal consistency was good (global Cronbach's alpha of 0.83 for Progress and 0.82 for obstruction). Measurement invariance was found across samples and gender, and the two-factor model obtained a good fit to the data. The latent means of progress and obstruction of the clinical sample were lower and higher, respectively, than the latent means of the nonclinical samples. Correlations with other variables were in the expected direction. CONCLUSION: The Spanish version of the VQ showed good psychometric properties.

A Hybrid Genome Assembly Resource for Podosphaera xanthii, the Main Causal Agent of Powdery Mildew Disease in Cucurbits.

Podosphaera xanthii is the main causal agent of powdery mildew in cucurbits and, arguably, the most important fungal pathogen of cucurbit crops. Here, we present the first reference genome assembly for P. xanthii. We performed a hybrid genome assembly, using reads from Illumina NextSeq550 and PacBio Sequel S3. The short and long reads were assembled into 1,727 scaffolds with an N50 size of 163,173 bp, resulting in a 142-Mb genome size. The combination of homology-based and ab initio predictions allowed the prediction of 14,911 complete genes. Repetitive sequences comprised 76.2% of the genome. Our P. xanthii genome assembly improves considerably the molecular resources for research on P. xanthii-cucurbit interactions and provides new opportunities for further genomics, transcriptomics, and evolutionary studies in powdery mildew fungi.[Formula: see text] Copyright © 2021 The Author(s). This is an open access article distributed under the CC BY-NC-ND 4.0 International license.

Cardiovascular mortality and cardiovascular event rates in patients with inflammatory rheumatic diseases in the CARdiovascular in rheuMAtology (CARMA) prospective study-results at 5 years of follow-up.

OBJECTIVES: To determine cardiovascular (CV) mortality and incidence of the first CV event (CVE) in patients with chronic inflammatory rheumatic diseases (CIRD) after 5 years of follow-up. METHODS: This is an analysis of the CARdiovascular in rheMAatology (CARMA) study after 5 years of follow-up. It includes patients with RA (n = 775), AS (n = 738) and PsA (n = 721), and individuals without CIRD (n = 677) attending outpatient rheumatology clinics from 67 public hospitals in Spain. Descriptive analyses were performed for the CV mortality at 5 years. The Systematic COronary Risk Evaluation (SCORE) function at 5 years was calculated to determine the expected risk of CV mortality. Poisson models were used to estimate the incidence rates of the first CVE. Hazard ratios of the risk factors involved in the development of the first CVE were evaluated using the Weibull proportional hazard model. RESULTS: Overall, 2382 subjects completed the follow-up visit at 5 years. Fifteen patients died due to CVE. CV deaths observed in the CIRD cohort were lower than that predicted by SCORE risk charts. The highest incidence rate of CVE [7.39 cases per 1000 person-years (95% CI 4.63, 11.18)] was found in PsA patients. However, after adjusting for age, sex and CV risk factors, AS was the inflammatory disease more commonly associated with CVE at 5 years [hazard ratio 4.60 (P =0.02)], compared with those without CIRD. CONCLUSIONS: Cardiovascular mortality in patients with CIRD at 5 years of follow-up is lower than estimated. Patients with AS have a higher risk of developing a first CVE after 5 years of follow-up.

Relevance of gastrointestinal manifestations in a large Spanish cohort of patients with systemic lupus erythematosus: what do we know?

OBJECTIVE: SLE can affect any part of the gastrointestinal (GI) tract. GI symptoms are reported to occur in >50% of SLE patients. To describe the GI manifestations of SLE in the RELESSER (Registry of SLE Patients of the Spanish Society of Rheumatology) cohort and to determine whether these are associated with a more severe disease, damage accrual and a worse prognosis. METHODS: We conducted a nationwide, retrospective, multicentre, cross-sectional cohort study of 3658 SLE patients who fulfil ≥4 ACR-97 criteria. Data on demographics, disease characteristics, activity (SLEDAI-2K or BILAG), damage (SLICC/ACR/DI) and therapies were collected. Demographic and clinical characteristics were compared between lupus patients with and without GI damage to establish whether GI damage is associated with a more severe disease. RESULTS: From 3654 lupus patients, 3.7% developed GI damage. Patients in this group (group 1) were older, they had longer disease duration, and were more likely to have vasculitis, renal disease and serositis than patients without GI damage (group 2). Hospitalizations and mortality were significantly higher in group 1. Patients in group 1 had higher modified SDI (SLICC Damage Index). The presence of oral ulcers reduced the risk of developing damage in 33% of patients. CONCLUSION: Having GI damage is associated with a worse prognosis. Patients on a high dose of glucocorticoids are at higher risk of developing GI damage which reinforces the strategy of minimizing glucocorticoids. Oral ulcers appear to decrease the risk of GI damage.

Parallel evolution of multiple mechanisms for demethylase inhibitor fungicide resistance in the barley pathogen Pyrenophora teres f. sp. maculata.

The fungal pathogen Pyrenophora teres f. sp. maculata (Ptm), responsible for spot-form of net blotch (SFNB), is currently the most significant disease of barley in Australia and a major disease worldwide. Management of SFNB relies heavily on fungicides and in Australia the demethylase inhibitors (DMIs) predominate. There have been sporadic reports of resistance to DMIs in Ptm but the mechanisms remain obscure. Ptm isolates collected from 1996 to 2019 in Western Australia were tested for fungicide sensitivity levels. Decreased sensitivity to DMIs was observed in isolates collected after 2015. Resistance factors to tebuconazole fell into two classes; moderate resistance (MR; RF 6-11) and high resistance (HR; RFs 30-65). Mutations linked to resistance were detected in the promoter region and coding sequence of the DMI target gene Cyp51A. Solo-LTR insertion elements were found at 5 different locations in the promoter region. Three different non-synonymous mutations encoded an altered protein with a phenylalanine to leucine substitution at position 489, F489L (F495I in the archetype CYP51A of Aspergillus fumigatus). F489L mutations have also been found in DMI-resistant strains of P. teres f. sp. teres. Ptm isolates carrying either a LTR insertion element or a F489L allele displayed the MR1 or MR2 phenotypes, respectively. Isolates carrying both an insertion element and a F489L mutation displayed the HR phenotype. Multiple mechanisms acting both alone and in concert were found to contribute to DMI resistance in Ptm. Moreover, these mutations have emerged repeatedly in Western Australian Ptm populations by a process of parallel evolution.

Prediction of pathogenicity genes involved in adaptation to a lupin host in the fungal pathogens Botrytis cinerea and Sclerotinia sclerotiorum via comparative genomics.

BACKGROUND: Narrow-leafed lupin is an emerging crop of significance in agriculture, livestock feed and human health food. However, its susceptibility to various diseases is a major obstacle towards increased adoption. Sclerotinia sclerotiorum and Botrytis cinerea - both necrotrophs with broad host-ranges - are reported among the top 10 lupin pathogens. Whole-genome sequencing and comparative genomics are useful tools to discover genes responsible for interactions between pathogens and their hosts. RESULTS: Genomes were assembled for one isolate of B. cinerea and two isolates of S. sclerotiorum, which were isolated from either narrow-leafed or pearl lupin species. Comparative genomics analysis between lupin-derived isolates and others isolated from alternate hosts was used to predict between 94 to 98 effector gene candidates from among their respective non-conserved gene contents. CONCLUSIONS: Detection of minor differences between relatively recently-diverged isolates, originating from distinct regions and with hosts, may highlight novel or recent gene mutations and losses resulting from host adaptation in broad host-range fungal pathogens.

Sampling for disease absence-deriving informed monitoring from epidemic traits.

Monitoring for disease requires subsets of the host population to be sampled and tested for the pathogen. If all the samples return healthy, what are the chances the disease was present but missed? In this paper, we developed a statistical approach to solve this problem considering the fundamental property of infectious diseases: their growing incidence in the host population. The model gives an estimate of the incidence probability density as a function of the sampling effort, and can be reversed to derive adequate monitoring patterns ensuring a given maximum incidence in the population. We then present an approximation of this model, providing a simple rule of thumb for practitioners. The approximation is shown to be accurate for a sample size larger than 20, and we demonstrate its use by applying it to three plant pathogens: citrus canker, bacterial blight and grey mould.

Increased PON lactonase activity in morbidly obese patients is associated with impaired lipid profile.

AIMS: The paraoxonase-1 (PON1) enzyme could play an important role in the anti-oxidant capacity of high-density lipoprotein. However, there are no studies which analyse the evolution of the three activities of PON1 (PON arylesterase, PON paraoxonase and PON lactonase) after Roux-en-Y Gastric Bypass (RYGB) in morbidly obese subjects. We analysed the association of PON concentration and activities with the evolution of morbidly obese subjects who underwent RYGB, and its relationship with biochemical variables and different atherogenic indices. METHODS: Twenty-seven non-obese and 82 morbidly obese subjects were studied before and 6 months after RYGB. RESULTS: Before RYGB, morbidly obese subjects had a lower PON1 concentration (P < 0.05) and higher PON lactonase activity (P < 0.001) than non-obese subjects, with no differences in PON arylesterase and PON paraoxonase activities. After RYGB, PON1 concentration (P < 0.05) and PON lactonase activity (P < 0.001) decreased with regard to the presurgery state. PON lactonase activity correlated with the atherogenic index of plasma before (r = 0.19, P = 0.047) and after RYGB (r = 0.27, P = 0.035). In different multiple lineal regression analysis models, presurgery PON lactonase activity was associated with total cholesterol (ß = 0.909, P < 0.001), LDL (ß = 0.632, P = 0.006) and DBP (ß = 0.230, P = 0.030) (R2  = 0.295), postsurgery PON lactonase activity was associated with esterified cholesterol (ß = 0.362, P = 0.011) (R2  = 0.131), and the change (Δ) in PON lactonase activity after RYGB was associated with Δesterified cholesterol (ß = 0.304, P = 0.030) (R2  = 0.093). CONCLUSIONS: PON lactonase activity is associated with the presence of morbid obesity and with an impaired lipid profile. All associations found could indicate the relationship between PON lactonase activity and the development of atherosclerosis.

Surveillance for azole resistance in clinical and environmental isolates of Aspergillus fumigatus in Australia and cyp51A homology modelling of azole-resistant isolates.

Background: The prevalence of azole resistance in Aspergillus fumigatus is uncertain in Australia. Azole exposure may select for resistance. We investigated the frequency of azole resistance in a large number of clinical and environmental isolates. Methods: A. fumigatus isolates [148 human, 21 animal and 185 environmental strains from air (n = 6) and azole-exposed (n = 64) or azole-naive (n = 115) environments] were screened for azole resistance using the VIPcheck™ system. MICs were determined using the Sensititre™ YeastOne YO10 assay. Sequencing of the Aspergillus cyp51A gene and promoter region was performed for azole-resistant isolates, and cyp51A homology protein modelling undertaken. Results: Non-WT MICs/MICs at the epidemiological cut-off value of one or more azoles were observed for 3/148 (2%) human isolates but not amongst animal, or environmental, isolates. All three isolates grew on at least one azole-supplemented well based on VIPcheck™ screening. For isolates 9 and 32, the itraconazole and posaconazole MICs were 1 mg/L (voriconazole MICs 0.12 mg/L); isolate 129 had itraconazole, posaconazole and voriconazole MICs of >16, 1 and 8 mg/L, respectively. Soil isolates from azole-exposed and azole-naive environments had similar geometric mean MICs of itraconazole, posaconazole and voriconazole (P > 0.05). A G54R mutation was identified in the isolates exhibiting itraconazole and posaconazole resistance, and the TR34/L98H mutation in the pan-azole-resistant isolate. cyp51A modelling predicted that the G54R mutation would prevent binding of itraconazole and posaconazole to the haem complex. Conclusions: Azole resistance is uncommon in Australian clinical and environmental A. fumigatus isolates; further surveillance is indicated.

Using Epidemiological Principles to Explain Fungicide Resistance Management Tactics: Why do Mixtures Outperform Alternations?

Whether fungicide resistance management is optimized by spraying chemicals with different modes of action as a mixture (i.e., simultaneously) or in alternation (i.e., sequentially) has been studied by experimenters and modelers for decades. However, results have been inconclusive. We use previously parameterized and validated mathematical models of wheat Septoria leaf blotch and grapevine powdery mildew to test which tactic provides better resistance management, using the total yield before resistance causes disease control to become economically ineffective ("lifetime yield") to measure effectiveness. We focus on tactics involving the combination of a low-risk and a high-risk fungicide, and the case in which resistance to the high-risk chemical is complete (i.e., in which there is no partial resistance). Lifetime yield is then optimized by spraying as much low-risk fungicide as is permitted, combined with slightly more high-risk fungicide than needed for acceptable initial disease control, applying these fungicides as a mixture. That mixture rather than alternation gives better performance is invariant to model parameterization and structure, as well as the pathosystem in question. However, if comparison focuses on other metrics, e.g., lifetime yield at full label dose, either mixture or alternation can be optimal. Our work shows how epidemiological principles can explain the evolution of fungicide resistance, and also highlights a theoretical framework to address the question of whether mixture or alternation provides better resistance management. It also demonstrates that precisely how spray tactics are compared must be given careful consideration. [Formula: see text] Copyright © 2018 The Author(s). This is an open access article distributed under the CC BY 4.0 International license .

The pro-/anti-inflammatory effects of different fatty acids on visceral adipocytes are partially mediated by GPR120.

PURPOSE: This study examines whether G-protein coupled receptor 120 (GPR120) is involved in the pro-/anti-inflammatory effects of different types of fatty acids (FAs) in human visceral adipocytes, and whether these effects may be altered in obesity, a state with a chronic inflammation. METHODS: Pro-/anti-inflammatory effects of palmitic, oleic, linoleic and docosahexaenoic acids on human visceral adipocytes were tested in mature adipocytes from non-obese and morbidly obese (MO) subjects. Also, the effects of these FAs were tested when the GPR120 gene was silenced. RESULTS: In adipocytes from non-obese subjects, palmitic and linoleic acids increased TNF-α and IL-6 mRNA expression (p < 0.05), and decreased IL-10 and adiponectin expression (p < 0.05). However, oleic and docosahexaenoic acids (DHA) produced the opposite effect (p < 0.05). In adipocytes from MO subjects, all FAs used increased TNF-α and IL-6 expression (p < 0.05). Palmitic and linoleic acids decreased IL-10 and adiponectin expression (p < 0.05), but oleic acid and DHA did not have significant effects. Only oleic acid increased adiponectin expression (p < 0.05). The effects of FAs on TNF-α, IL-6, IL-10 and adiponectin expression in non-obese and MO subjects were significantly annulled when the GPR120 gene was silenced in visceral adipocytes differentiated from human mesenchymal stem cells. CONCLUSIONS: FAs are capable of directly acting on visceral adipocytes to modulate differently TNF-α, IL-6, IL-10 and adiponectin expression, with a different and greater effect in MO subjects. These effects are largely annulled when GPR120 expression was silenced, which suggests that they could be mediated by GPR120.

Multifunctional decoration of alpha-tocopheryl succinate-based NP for cancer treatment: effect of TPP and LTVSPWY peptide.

Active targeting not only of a specific cell but also a specific organelle maximizes the therapeutic activity minimizing adverse side effects in healthy tissues. The present work describes the synthesis, characterization, and in vitro biological activity of active targeting nanoparticles (NP) for cancer therapy based on α-tocopheryl succinate (α-TOS), a well-known mitocan, that selectively induces apoptosis of cancer cells and proliferalting endothelial cells. Human epidermal growth factor receptor 2 (HER2) targeting peptide LTVSPWY (PEP) and triphenylphosphonium lipophilic cation (TPP) were conjugated to a previously optimized RAFT block copolymer that formed self-assembled NP of appropriate size for this application and low polydispersity by self-organized precipitation method. PEP and TPP were included in order to target not only HER2 positive cancer cells, but also the mitochondria of these cancer cells, respectively. The in vitro experiments demonstrated the faster incorporation of the active-targeting NP and the higher accumulation of TPP-bearing NP in the mitochondria of MDA-MB-453 HER2 positive cancer cells compared to non-decorated NP. Moreover, the encapsulation of additional α-TOS in the hydrophobic core of the NP was achieved with high efficiencies. The loaded NP presented higher cytotoxicity than unloaded NP but preserved their selectivity against cancer cells in a range of tested concentrations.

Olive Actual "on Year" Yield Forecast Tool Based on the Tree Canopy Geometry Using UAS Imagery.

Olive has a notable importance in countries of Mediterranean basin and its profitability depends on several factors such as actual yield, production cost or product price. Actual "on year" Yield (AY) is production (kg tree-1) in "on years", and this research attempts to relate it with geometrical parameters of the tree canopy. Regression equation to forecast AY based on manual canopy volume was determined based on data acquired from different orchard categories and cultivars during different harvesting seasons in southern Spain. Orthoimages were acquired with unmanned aerial systems (UAS) imagery calculating individual crown for relating to canopy volume and AY. Yield levels did not vary between orchard categories; however, it did between irrigated orchards (7000-17,000 kg ha-1) and rainfed ones (4000-7000 kg ha-1). After that, manual canopy volume was related with the individual crown area of trees that were calculated by orthoimages acquired with UAS imagery. Finally, AY was forecasted using both manual canopy volume and individual tree crown area as main factors for olive productivity. AY forecast only by using individual crown area made it possible to get a simple and cheap forecast tool for a wide range of olive orchards. Finally, the acquired information was introduced in a thematic map describing spatial AY variability obtained from orthoimage analysis that may be a powerful tool for farmers, insurance systems, market forecasts or to detect agronomical problems.

Infinite Continuous Feature Model for Psychiatric Comorbidity Analysis.

We aim at finding the comorbidity patterns of substance abuse, mood and personality disorders using the diagnoses from the National Epidemiologic Survey on Alcohol and Related Conditions database. To this end, we propose a novel Bayesian nonparametric latent feature model for categorical observations, based on the Indian buffet process, in which the latent variables can take values between 0 and 1. The proposed model has several interesting features for modeling psychiatric disorders. First, the latent features might be off, which allows distinguishing between the subjects who suffer a condition and those who do not. Second, the active latent features take positive values, which allows modeling the extent to which the patient has that condition. We also develop a new Markov chain Monte Carlo inference algorithm for our model that makes use of a nested expectation propagation procedure.

Olive Crown Porosity Measurement Based on Radiation Transmittance: An Assessment of Pruning Effect.

Crown porosity influences radiation interception, air movement through the fruit orchard, spray penetration, and harvesting operation in fruit crops. The aim of the present study was to develop an accurate and reliable methodology based on transmitted radiation measurements to assess the porosity of traditional olive trees under different pruning treatments. Transmitted radiation was employed as an indirect method to measure crown porosity in two olive orchards of the Picual and Hojiblanca cultivars. Additionally, three different pruning treatments were considered to determine if the pruning system influences crown porosity. This study evaluated the accuracy and repeatability of four algorithms in measuring crown porosity under different solar zenith angles. From a 14° to 30° solar zenith angle, the selected algorithm produced an absolute error of less than 5% and a repeatability higher than 0.9. The described method and selected algorithm proved satisfactory in field results, making it possible to measure crown porosity at different solar zenith angles. However, pruning fresh weight did not show any relationship with crown porosity due to the great differences between removed branches. A robust and accurate algorithm was selected for crown porosity measurements in traditional olive trees, making it possible to discern between different pruning treatments.