A mixture of delta-rules approximation to bayesian inference in change-point problems

Error-driven learning rules have received considerable attention because of their close relationships to both optimal theory and neurobiological mechanisms. However, basic forms of these rules are effective under only a restricted set of conditions in which the environment is stable. Recent studies have defined optimal solutions to learning problems in more general, potentially unstable, environments, but the relevance of these complex mathematical solutions to how the brain solves these problems remains unclear. Here, we show that one such Bayesian solution can be approximated by a computationally straightforward mixture of simple error-driven 'Delta' rules. This simpler model can make effective inferences in a dynamic environment and matches human performance on a predictive-inference task using a mixture of a small number of Delta rules. This model represents an important conceptual advance in our understanding of how the brain can use relatively simple computations to make nearly optimal inferences in a dynamic world.

Dehydrin, alcohol dehydrogenase, and central metabolite levels are associated with cold tolerance in diploid strawberry (Fragaria spp.)

The use of artificial freezing tests, identification of biomarkers linked to or directly involved in the low-temperature tolerance processes, could prove useful in applied strawberry breeding. This study was conducted to identify genotypes of diploid strawberry that differ in their tolerance to low-temperature stress and to investigate whether a set of candidate proteins and metabolites correlate with the level of tolerance. 17 Fragaria vesca, 2 F. nilgerrensis, 2 F. nubicola, and 1 F. pentaphylla genotypes were evaluated for low-temperature tolerance. Estimates of temperatures where 50 % of the plants survived (LT₅₀) ranged from -4.7 to -12.0 °C between the genotypes. Among the F. vesca genotypes, the LT₅₀ varied from -7.7 °C to -12.0 °C. Among the most tolerant were three F. vesca ssp. bracteata genotypes (FDP821, NCGR424, and NCGR502), while a F. vesca ssp. californica genotype (FDP817) was the least tolerant (LT₅₀) -7.7 °C). Alcohol dehydrogenase (ADH), total dehydrin expression, and content of central metabolism constituents were assayed in select plants acclimated at 2 °C. The LT₅₀ estimates and the expression of ADH and total dehydrins were highly correlated (r(adh) = -0.87, r (dehyd) = -0.82). Compounds related to the citric acid cycle were quantified in the leaves during acclimation. While several sugars and acids were significantly correlated to the LT₅₀ estimates early in the acclimation period, only galactinol proved to be a good LT₅₀ predictor after 28 days of acclimation (r(galact) = 0.79). It is concluded that ADH, dehydrins, and galactinol show great potential to serve as biomarkers for cold tolerance in diploid strawberry.

Proteomic study of low-temperature responses in strawberry cultivars (Fragaria x ananassa) that differ in cold tolerance

To gain insight into the molecular basis contributing to overwintering hardiness, a comprehensive proteomic analysis comparing crowns of octoploid strawberry (Fragaria × ananassa) cultivars that differ in freezing tolerance was conducted. Four cultivars were examined for freeze tolerance and the most cold-tolerant cultivar ('Jonsok') and least-tolerant cultivar ('Frida') were compared with a goal to reveal how freezing tolerance is achieved in this distinctive overwintering structure and to identify potential cold-tolerance-associated biomarkers. Supported by univariate and multivariate analysis, a total of 63 spots from two-dimensional electrophoresis analysis and 135 proteins from label-free quantitative proteomics were identified as significantly differentially expressed in crown tissue from the two strawberry cultivars exposed to 0-, 2-, and 42-d cold treatment. Proteins identified as cold-tolerance-associated included molecular chaperones, antioxidants/detoxifying enzymes, metabolic enzymes, pathogenesis-related proteins, and flavonoid pathway proteins. A number of proteins were newly identified as associated with cold tolerance. Distinctive mechanisms for cold tolerance were characterized for two cultivars. In particular, the 'Frida' cold response emphasized proteins specific to flavonoid biosynthesis, while the more freezing-tolerant 'Jonsok' had a more comprehensive suite of known stress-responsive proteins including those involved in antioxidation, detoxification, and disease resistance. The molecular basis for 'Jonsok'-enhanced cold tolerance can be explained by the constitutive level of a number of proteins that provide a physiological stress-tolerant poise.

Rational regulation of learning dynamics by pupil-linked arousal systems

The ability to make inferences about the current state of a dynamic process requires ongoing assessments of the stability and reliability of data generated by that process. We found that these assessments, as defined by a normative model, were reflected in nonluminance-mediated changes in pupil diameter of human subjects performing a predictive-inference task. Brief changes in pupil diameter reflected assessed instabilities in a process that generated noisy data. Baseline pupil diameter reflected the reliability with which recent data indicate the current state of the data-generating process and individual differences in expectations about the rate of instabilities. Together these pupil metrics predicted the influence of new data on subsequent inferences. Moreover, a task- and luminance-independent manipulation of pupil diameter predictably altered the influence of new data. Thus, pupil-linked arousal systems can help to regulate the influence of incoming data on existing beliefs in a dynamic environment.

Metabolite profiling reveals novel multi-level cold responses in the diploid model Fragaria vesca (woodland strawberry)

Winter freezing damage is a crucial factor in overwintering crops such as the octoploid strawberry (Fragaria × ananassa Duch.) when grown in a perennial cultivation system. Our study aimed at assessing metabolic processes and regulatory mechanisms in the close-related diploid model woodland strawberry (Fragaria vescaL.) during a 10-days cold acclimation experiment. Based on gas chromatography/time-of-flight-mass spectrometry (GC/TOF-MS) metabolite profiling of three F. vesca genotypes, clear distinctions could be made between leaves and non-photosynthesizing roots, underscoring the evolvement of organ-dependent cold acclimation strategies. Carbohydrate and amino acid metabolism, photosynthetic acclimation, and antioxidant and detoxification systems (ascorbate pathway) were strongly affected. Metabolic changes in F. vesca included the strong modulation of central metabolism, and induction of osmotically-active sugars (fructose, glucose), amino acids (aspartic acid), and amines (putrescine). In contrast, a distinct impact on the amino acid proline, known to be cold-induced in other plant systems, was conspicuously absent. Levels of galactinol and raffinose, key metabolites of the cold-inducible raffinose pathway, were drastically enhanced in both leaves and roots throughout the cold acclimation period of 10 days. Furthermore, initial freezing tests and multifaceted GC/TOF-MS data processing (Venn diagrams, independent component analysis, hierarchical clustering) showed that changes in metabolite pools of cold-acclimated F. vesca were clearly influenced by genotype.

Inferring relevance in a changing world

Reinforcement learning models of human and animal learning usually concentrate on how we learn the relationship between different stimuli or actions and rewards. However, in real-world situations “stimuli” are ill-defined. On the one hand, our immediate environment is extremely multidimensional. On the other hand, in every decision making scenario only a few aspects of the environment are relevant for obtaining reward, while most are irrelevant. Thus a key question is how do we learn these relevant dimensions, that is, how do we learn what to learn about? We investigated this process of “representation learning” experimentally, using a task in which one stimulus dimension was relevant for determining reward at each point in time. As in real life situations, in our task the relevant dimension can change without warning, adding ever-present uncertainty engendered by a constantly changing environment. We show that human performance on this task is better described by a suboptimal strategy based on selective attention and serial-hypothesis-testing rather than a normative strategy based on probabilistic inference. From this, we conjecture that the problem of inferring relevance in general scenarios is too computationally demanding for the brain to solve optimally. As a result the brain utilizes approximations, employing these even in simplified scenarios in which optimal representation learning is tractable, such as the one in our experiment.

Isolated mild clitoral hypertrophy may reveal 46,XY disorders of sex development in infancy due to 17βHSD-3 defect confirmed by molecular analysis

AIMS

17-β-Hydroxysteroid dehydrogenase type 3 (17βHSD-3) is expressed exclusively in the testes where it converts Δ4 androstenedione (Δ4) to testosterone (T). Here, we report a patient with a rare mutation at a critical site in HSD17B3 gene leading to deficiency of 17β HSD-3 enzyme.

METHODS

We describe a 3-year old healthy female of consanguineous Lebanese descent, who presented to the endocrine service with isolated mild clitoromegaly. Adrenocorticotropic hormone (ACTH) and human chorionic gonadotrophin (hCG) stimulation tests were performed. Genes for sex-determining region Y (SRY), steroidogenic factor-1 (SF-1) and 17βHSD-3 (HSD17B3) were sequenced.

RESULTS

The post-hCG stimulation T levels and T/Δ4 ratio was low. Patient had a 46,XY karyotype. Sequence analysis of the HSD17B3 gene revealed a homozygous R80W missense mutation on exon 3. No mutation was found in SRY and SF1 genes. Mullerian structures were not detected on pelvic imaging.

CONCLUSIONS

A low T/Δ4 ratio is indicative of 17βHSD-3 deficiency and associated with isolated clitoromegaly. The R80 site is critical for NADPH binding, thus the mutation at this site leads to 17βHSD-3 deficiency presenting as 46,XY disorder of sex development.

The restructuring of institutional long-term care in St. John's: impact of supply-induced demand on planning

Restructuring of institutional long-term care was undertaken using predictions of future bed need with assumptions made on incidence rates of clients defined by type of disability, survival, and demographic changes. Recent substantial increase in the population rate of clients seeking placement across all degrees of disability, coincident with new facilities for those with modest disability, occurred. Consequently, more appropriate housing and supervised care beds, and more limited downsizing of nursing homes will be required.

Bayesian online learning of the hazard rate in change-point problems

Change-point models are generative models of time-varying data in which the underlying generative parameters undergo discontinuous changes at different points in time known as change points. Change-points often represent important events in the underlying processes, like a change in brain state reflected in EEG data or a change in the value of a company reflected in its stock price. However, change-points can be difficult to identify in noisy data streams. Previous attempts to identify change-points online using Bayesian inference relied on specifying in advance the rate at which they occur, called the hazard rate (h). This approach leads to predictions that can depend strongly on the choice of h and is unable to deal optimally with systems in which h is not constant in time. In this letter, we overcome these limitations by developing a hierarchical extension to earlier models. This approach allows h itself to be inferred from the data, which in turn helps to identify when change-points occur. We show that our approach can effectively identify change-points in both toy and real data sets with complex hazard rates and how it can be used as an ideal-observer model for human and animal behavior when faced with rapidly changing inputs.

Genomic organization and gene expression of the multiple globins in Atlantic cod: conservation of globin-flanking genes in chordates infers the origin of the vertebrate globin clusters

Background

The vertebrate globin genes encoding the α- and β-subunits of the tetrameric hemoglobins are clustered at two unlinked loci. The highly conserved linear order of the genes flanking the hemoglobins provides a strong anchor for inferring common ancestry of the globin clusters. In fish, the number of α-β-linked globin genes varies considerably between different sublineages and seems to be related to prevailing physico-chemical conditions. Draft sequences of the Atlantic cod genome enabled us to determine the genomic organization of the globin repertoire in this marine species that copes with fluctuating environments of the temperate and Arctic regions.

Results

The Atlantic cod genome was shown to contain 14 globin genes, including nine hemoglobin genes organized in two unlinked clusters designated β5-α1-β1-α4 and β3-β4-α2-α3-β2. The diverged cod hemoglobin genes displayed different expression levels in adult fish, and tetrameric hemoglobins with or without a Root effect were predicted. The novel finding of maternally inherited hemoglobin mRNAs is consistent with a potential role played by fish hemoglobins in the non-specific immune response. In silico analysis of the six teleost genomes available showed that the two α-β globin clusters are flanked by paralogs of five duplicated genes, in agreement with the proposed teleost-specific duplication of the ancestral vertebrate globin cluster. Screening the genome of extant urochordate and cephalochordate species for conserved globin-flanking genes revealed linkage of RHBDF1, MPG and ARHGAP17 to globin genes in the tunicate Ciona intestinalis, while these genes together with LCMT are closely positioned in amphioxus (Branchiostoma floridae), but seem to be unlinked to the multiple globin genes identified in this species.

Conclusion

The plasticity of Atlantic cod to variable environmental conditions probably involves the expression of multiple globins with potentially different properties. The interspecific difference in number of fish hemoglobin genes contrasts with the highly conserved synteny of the flanking genes. The proximity of globin-flanking genes in the tunicate and amphioxus genomes resembles the RHBDF1-MPG-α-globin-ARHGAP17-LCMT linked genes in man and chicken. We hypothesize that the fusion of the three chordate linkage groups 3, 15 and 17 more than 800 MYA led to the ancestral vertebrate globin cluster during a geological period of increased atmospheric oxygen content.

Real-time landslide warning during heavy rainfall

A real-time system for issuing warnings of landslides during major storms is being developed for the San Francisco Bay region, California. The system is based on empirical and theoretical relations between rainfall and landslide initiation, geologic determination of areas susceptible to landslides, real-time monitoring of a regional network of telemetering rain gages, and National Weather Service precipitation forecasts. This system was used to issue warnings during the storms of 12 to 21 February 1986, which produced 800 millimeters of rainfall in the region. Although analysis after the storms suggests that modifications and additional development are needed, the system successfully predicted the times of major landslide events. It could be used as a prototype for systems in other landslide-prone regions.

An approach for the assessment of risk from chronic radiation to populations of phytoplankton and zooplankton

A conceptual model of the effects of chronic radiation on a population of phytoplankton and zooplankton in an oceanic nutrient layer is presented. The model shows that there are distinct threshold dose rates at which the different plankton populations become unsustainable. These are 10,400 microGy h(-1) for phytoplankton and 125 microGy h(-1) for zooplankton. Both these values are considerably greater than the current screening values for protection of 10 microGy h(-1). The model highlights the effects of predator-prey dynamics in predicting that when the zooplankton is affected by the radiation dose, the phytoplankton population can increase. In addition, the model was altered to replicate the dose rates to the plankton of a previous ERICA Irish Sea assessment (24 microGy h(-1) for zooplankton and 430 microGy h(-1) to phytoplankton). The results showed only a 10% decrease in the zooplankton population and a 15% increase in the phytoplankton population. Therefore, at this level of dose, the model predicts that although the dose rate exceeds the guideline value, populations are not significantly affected. This result highlights the limitations of a single screening value for different groups of organisms.

An approach to the assessment of risk from chronic radiation to populations of European lobster, Homarus gammarus (L.)

The basic principles underlying a four-discrete age group, logistic, growth model for the European lobster Homarus gammarus are presented and discussed at proof-of-concept level. The model considers reproduction, removal by predation, natural death, fishing, radiation and migration. Non-stochastic effects of chronic low linear energy transfer (LET) radiation are modelled with emphasis on (99)Tc, using three endpoints: repairable radiation damage, impairment of reproductive ability and, at higher dose rates, mortality. An allometric approach for the calculation of LD(50/30) as a function of the mass of each life stage is used in model calibration. The model predicts that at a dose rate of 1 Gy day(-1), lobster population reproduction and survival become severely compromised, leading eventually to population extinction. At 0.01 Gy day(-1), the survival rate of an isolated population is reduced by 10%, mainly through loss of fecundity, comparable to natural migration losses. Fishing is the main ecological stress and only dose rates in the range 0.03-0.1 Gy day(-1) can achieve discernible effects above it. On the balance of radiation and other ecological stresses, a benchmark value of 0.01 Gy day(-1) is proposed for the protection of lobster populations. This value appears consistent with available information on radiation effects in wildlife.

Laboratory and field studies of polonium and plutonium in marine plankton

Uptake experiments were successfully undertaken for polonium and plutonium in the marine diatom Skeletonema costatum. Experiments were complemented with a field study which yielded concentration factors for these radionuclides for plankton samples collected in the eastern Irish Sea. The uptake experiment produced uptake half times of 0.8 and 0.6 days for plutonium and polonium, respectively. Concentration factors of 3.1 x 10(5) l kg(-1) for plutonium and 2.4 x 10(5) l kg(-1) for polonium were consistent with those suggested by the IAEA for phytoplankton. Concentration factors derived from the field study were 2 x 10(3) l kg(-1) for plutonium and 1.4 x 10(4) l kg(-1) for polonium and were more consistent with IAEA values for zooplankton rather than phytoplankton. The maximum calculated dose rate due to Po and Pu in plankton sampled in the field was 0.12 microGyh(-1). In applying the laboratory-derived phytoplankton CF values for Po and Pu, higher dose rates, in the range 6-11 microGyh(-1), were calculated. The latter values are well below the UNSCEAR and IAEA review value of 400 microGyh(-1), and are also below the Environment Agency marine action level of 40 microGyh(-1).

Parallel Hopfield networks

We introduce a novel type of neural network, termed the parallel Hopfield network, that can simultaneously effect the dynamics of many different, independent Hopfield networks in parallel in the same piece of neural hardware. Numerically we find that under certain conditions, each Hopfield subnetwork has a finite memory capacity approaching that of the equivalent isolated attractor network, while a simple signal-to-noise analysis sheds qualitative, and some quantitative, insight into the workings (and failures) of the system.

CYP21A2 gene mutations in congenital adrenal hyperplasia: genotype-phenotype correlation in Turkish children

BACKGROUND

Congenital adrenal hyperplasia (CAH) due 21-hydroxylase deficiency (21-OHD) is a common autosomal recessive disorder. It is caused by defects in the CYP21A2 gene.

OBJECTIVE

Our aim was to determine the frequency of common gene mutations and to evaluate genotype-phenotype correlations in Turkish 21-OHD patients.

METHODS

Molecular analysis of the CYP21A2 gene was performed for the detection of the eight most common point mutations [p.P30L, IVS2-13C>G (IVS-2), p.I172N, exon 6 mutation cluster (p.I236N, p.V237E, p.M239K), p.V281L, p.Q318X, p.R356W, 8-bp-deletion], of large deletion and conversion by southern blotting, allele specific semi-quantitative PCR/enzyme restriction method and sequencing, in 56 patients with 21-OHD, from 52 families.

RESULTS

Disease-causing mutations were identified in 77 out of 91 alleles (84.6%) of the patients. Mutations were found in 34 of 43 alleles (79.1%) in salt wasting (SW; n=26), 32 of 36 alleles (88.8%) in simple virilizing (SV; n=24) and 11 of 12 alleles (91.6%) in non-classical (NC; n=6) form of CAH. The most frequent mutations were IVS-2 (22.0%), large conversion (14.3%), p.I172N (9.9%) p.R356W (8.8%), and large deletion (6.6%). In the SW form, the most frequent genotypes were homozygous for IVS-2 (11.5%) and homozygous for large conversion of the gene (11.5%). In the SV form, the most frequent genotype was homozygous for IVS-2 (20%), followed by compound heterozygous for p.I172N/8-bp del (10%). Homozygous for p.V281L (16.7%) was most common in NC. In most cases there was good correlation between genotype and phenotype. In the SW and NC forms, genotypes of all the patients correlated with their phenotypes.

CONCLUSIONS

This is the first comprehensive study on the molecular basis of CAH patients in the Turkish population. Based on these results, we propose a modified screening strategy to facilitate molecular testing of CAH patients in our population.

Dynamic model for the assessment of radiological exposure to marine biota

A generic approach has been developed to simulate dynamically the uptake and turnover of radionuclides by marine biota. The approach incorporates a three-compartment biokinetic model based on first order linear kinetics, with interchange rates between the organism and its surrounding environment. Model rate constants are deduced as a function of known parameters: biological half-lives of elimination, concentration factors and a sample point of the retention curve, allowing for the representation of multi-component release. The new methodology has been tested and validated in respect of non-dynamic assessment models developed for regulatory purposes. The approach has also been successfully tested against research dynamic models developed to represent the uptake of technetium and radioiodine by lobsters and winkles. Assessments conducted on two realistic test scenarios demonstrated the importance of simulating time-dependency for ecosystems in which environmental levels of radionuclides are not in equilibrium.

Viral suppressor proteins show varying abilities and effectiveness to suppress transgene-induced post-transcriptional gene silencing of endogenous Chalcone synthase in transgenic Arabidopsis

Many, if not most, plant viruses encode proteins that interfere with RNA silencing pathways in plants. These proteins, known as viral suppressor proteins interfere at different key steps of the silencing pathways, and are able to suppress, to varying degree, transgene-induced silencing in plants. In this study, we report the ability and effectiveness of four different viral suppressor proteins that interfere with post-transcriptional gene silencing (PTGS) of the endogenous chalcone synthase gene (CHS) in Arabidopsis when the silencing trigger and the viral suppressor protein were expressed from the same transgene locus. The silencing trigger consisted of an inverted-repeat transgene construct that induces PTGS of the endogenous Arabidopsis CHS gene with high efficiency. Real-time PCR analyses were used to monitor the transcript levels of both the viral mRNAs and the target CHS mRNAs in transgenic Arabidopsis. Our results show that only one of the viral suppressor proteins tested, the p38 protein of Turnip Crinkle Virus, was able to efficiently, albeit to varying degrees, interfere with PTGS of CHS in individual transgenic Arabidopsis plants. Moreover, we show that the degree of CHS silencing suppression was dependent on the transcript level of p38. In contrast to earlier reports, we also found that constitutive expression of p38 in transgenic Arabidopsis is correlated with morphological defects in leaves and flowers.

A study of pattern recovery in recurrent correlation associative memories

In this paper, we analyze the recurrent correlation associative memory (RCAM) model of Chiueh and Goodman (1990, 1991). This is an associative memory in which stored binary memory patterns are recalled via an iterative update rule. The update of the individual pattern-bits is controlled by an excitation function, which takes as its argument the inner product between the stored memory patterns and the input patterns. Our contribution is to analyze the dynamics of pattern recall when the input patterns are corrupted by noise of a relatively unrestricted class. We show how to identify the excitation function which maximizes the separation (the Fisher discriminant) between the uncorrupted realization of the noisy input pattern and the remaining patterns residing in the memory. The excitation function which gives maximum separation is exponential when the input bit-errors follow a binomial distribution. We develop an expression for the expectation value of bit-error probability on the input pattern after one iteration. We show how to identify the excitation function which minimizes the bit-error probability. The relationship between the excitation functions which result from the two different approaches is examined for a binomial distribution of bit-errors. We develop a semiempirical approach to the modeling of the dynamics of the RCAM.