Structure–property correlation over five phases and four transitions in Pb5Al3F19

The calorimetric and dielectric properties of Pb5Al3F19 in the five phases stable under ambient pressure are correlated with structure for fuller characterization of each phase. The first-order transition between ferroelectric phase V and antiferroelectric phase IV at T V,IV = 260 (5) K exhibits a thermal hysteresis of 135 (5) K on heating, with a maximum atomic displacement Δ(xyz)max = 1.21 (6) Å; the transition from phase IV to ferroelastic phase III at 315 (5) K is also first order but with a thermal hysteresis of 10 (5) K and Δ(xyz)max = 0.92 (7)  Å; that from phase III to paraelastic phase II at 360 (5) K is second order without hysteresis and has Δ(xyz)max = 0.69 (4) Å; and the transition from phase II to paraelectric phase I at 670 (5) K is second or higher order, with Δ(xyz)max = 0.7 (4) Å. The measured entropy change ΔS at T V,IV agrees well with ΔS as derived from the increased configurational energy by Stirling's approximation. For all other phase transitions, 0.5 ≥ ΔS > 0 J mol−1 K−1 is consistent with an entropy change caused primarily by the changes in the vibrational energy. The structure of phase III is determined both by group theoretical/normal mode analysis and by consideration of the structures of phases II, IV and V reported previously; refinement is by simultaneous Rietveld analysis of the X-ray and neutron diffraction powder profiles. The structure of prototypic phase I is predicted on the basis of the atomic arrangement in phases II, III, IV and V. The introduction of 3d electrons into the Pb5Al3F19 lattice disturbs the structural equilibrium, the addition of 0.04% Cr3+ causing significant changes in atomic positions and increasing T IV,III by ∼15 K. Substitution of Al3+ by 20% or more Cr3+ eliminates the potential minima that otherwise stabilize phases IV, III and II.

Neural architectures for robot intelligence

We argue that direct experimental approaches to elucidate the architecture of higher brains may benefit from insights gained from exploring the possibilities and limits of artificial control architectures for robot systems. We present some of our recent work that has been motivated by that view and that is centered around the study of various aspects of hand actions since these are intimately linked with many higher cognitive abilities. As examples, we report on the development of a modular system for the recognition of continuous hand postures based on neural nets, the use of vision and tactile sensing for guiding prehensile movements of a multifingered hand, and the recognition and use of hand gestures for robot teaching. Regarding the issue of learning, we propose to view real-world learning from the perspective of data-mining and to focus more strongly on the imitation of observed actions instead of purely reinforcement-based exploration. As a concrete example of such an effort we report on the status of an ongoing project in our laboratory in which a robot equipped with an attention system with a neurally inspired architecture is taught actions by using hand gestures in conjunction with speech commands. We point out some of the lessons learnt from this system, and discuss how systems of this kind can contribute to the study of issues at the junction between natural and artificial cognitive systems.

Chromosome 18 replaced by two ring chromosomes of chromosome 18 origin

We here describe the first example of the replacement of an autosome by two ring chromosomes originating from the missing chromosome, presented in a patient with a single chromosome 18 and two additional ring chromosomes. Detailed fluorescence in situ hybridization (FISH) analysis revealed the chromosome 18 origin of both ring chromosomes and characterized the small and the large ring chromosome as derivatives of the short and long arm of chromosome 18, respectively. The loss of subtelomeric regions of the short and the long arm of chromosome 18 in the ring chromosomes was confirmed by FISH studies. Molecular studies showed the exclusive presence of the paternal alleles for microsatellite markers located distal to the short and long arm loci D18S843 and D18S474, respectively. This indicates the maternal origin of both rings and provides evidence for substantial deletions of the distal parts of both arms of chromosome 18 in the ring chromosomes. The dysmorphic features of the patient can be explained by these deletions in both chromosome arms, as the clinical findings partly overlap with observations in 18p- and 18q-syndrome and are similar to some cases of ring chromosome 18. Centromere misdivision is suggested as one mechanism involved in the formation of the ring chromosomes.

Automatic Recognition of Muscle-Invasive T-Lymphocytes Expressing Dipeptidyl-Peptidase IV (CD26) and Analysis of the Associated Cell Surface Phenotypes

A neural cell detection system (NCDS) for the automatic quantitation of fluorescent lymphocytes in tissue sections was used to analyze CD26 expression in muscle-invasive T-cells. CD26 is a cell surface dipeptidyl-peptidase IV (DPP IV) involved in co-stimulatory activation of T-cells and also in adhesive events. The NCDS system acquires visual knowledge from a set of training cell image patches selected by a user. The trained system evaluates an image in 2 min calculating (i) the number, (ii) the positions and (iii) the phenotypes of the fluorescent cells. In the present study we have used the NCDS to identity DPP IV (CD26) expressing invasive lymphocytes in sarcoid myopathy and to analyze the associated cell surface phenotypes. We find highly unusual phenotypes characterized by differential combination of seven cell surface receptors usually involved in co-stimulatory events in T-lymphocytes. The data support a differential adhesive rather than a co-stimulatory role of CD26 in muscle-invasive cells. The adaptability of the NCDS algorithm to diverse types of cells should enable us to approach any invasion process, including invasion of malignant cells.

Dynamical stability conditions for recurrent neural networks with unsaturating piecewise linear transfer functions

We establish two conditions that ensure the nondivergence of additive recurrent networks with unsaturating piecewise linear transfer functions, also called linear threshold or semilinear transfer functions. As Hahnloser, Sarpeshkar, Mahowald, Douglas, and Seung (2000) showed, networks of this type can be efficiently built in silicon and exhibit the coexistence of digital selection and analog amplification in a single circuit. To obtain this behavior, the network must be multistable and nondivergent, and our conditions allow determining the regimes where this can be achieved with maximal recurrent amplification. The first condition can be applied to nonsymmetric networks and has a simple interpretation of requiring that the strength of local inhibition match the sum over excitatory weights converging onto a neuron. The second condition is restricted to symmetric networks, but can also take into account the stabilizing effect of nonlocal inhibitory interactions. We demonstrate the application of the conditions on a simple example and the orientation-selectivity model of Ben-Yishai, Lev Bar-Or, and Sompolinsky (1995). We show that the conditions can be used to identify in their model regions of maximal orientation-selective amplification and symmetry breaking.

Heterozygous mutations in ANKH, the human ortholog of the mouse progressive ankylosis gene, result in craniometaphyseal dysplasia

Craniometaphyseal dysplasia (CMD) is a bone dysplasia characterized by overgrowth and sclerosis of the craniofacial bones and abnormal modeling of the metaphyses of the tubular bones. Hyperostosis and sclerosis of the skull may lead to cranial nerve compressions resulting in hearing loss and facial palsy. An autosomal dominant form of the disorder (MIM 123000) was linked to chromosome 5p15.2-p14.1 (ref. 3) within a region harboring the human homolog (ANKH) of the mouse progressive ankylosis (ank) gene. The ANK protein spans the outer cell membrane and shuttles inorganic pyrophosphate (PPi), a major inhibitor of physiologic and pathologic calcification, bone mineralization and bone resorption. Here we carry out mutation analysis of ANKH, revealing six different mutations in eight of nine families. The mutations predict single amino acid substitutions, deletions or insertions. Using a helix prediction program, we propose for the ANK molecule 12 membrane-spanning helices with an alternate inside/out orientation and a central channel permitting the passage of PPi. The mutations occur at highly conserved amino acid residues presumed to be located in the cytosolic portion of the protein. Our results link the PPi channel ANK with bone formation and remodeling.

A competitive-layer model for feature binding and sensory segmentation

We present a recurrent neural network for feature binding and sensory segmentation: the competitive-layer model (CLM). The CLM uses topographically structured competitive and cooperative interactions in a layered network to partition a set of input features into salient groups. The dynamics is formulated within a standard additive recurrent network with linear threshold neurons. Contextual relations among features are coded by pairwise compatibilities, which define an energy function to be minimized by the neural dynamics. Due to the usage of dynamical winner-take-all circuits, the model gains more flexible response properties than spin models of segmentation by exploiting amplitude information in the grouping process. We prove analytic results on the convergence and stable attractors of the CLM, which generalize earlier results on winner-take-all networks, and incorporate deterministic annealing for robustness against local minima. The piecewise linear dynamics of the CLM allows a linear eigensubspace analysis, which we use to analyze the dynamics of binding in conjunction with annealing. For the example of contour detection, we show how the CLM can integrate figure-ground segmentation and grouping into a unified model.

The short tandem repeat loci hTPO, THO1 and FGA

This paper focuses on the three well-established short tandem repeats HumhTPO (human thyroid peroxidase gene), HumTHO1 (human tyrosine hydroxylase gene) and HumFGA (human alpha fibrinogen gene). We present sequencing data which include the sequences of all the different primer pairs used so far, as well as results from a population study which was carried out on an American Black and a southwestern German population. All data are compared with previously published studies. A test for homogeneity confirmed that non-European populations possess different genetic pools for each of the three loci under investigation.

The short tandem repeat locus D3S1359

Short tandem repeats (STRs) are used by many laboratories throughout the world performing paternity testing or criminal casework. Nevertheless, many of the established STRs have obvious disadvantages such as low number of common alleles (e.g., hTPO, THO1) or alleles with frequencies of nearly 50% (e.g., hTPO, FES). In this paper the new STR locus D3S1359 is described. In a population study which was carried out on 136 unrelated individuals from southwestern Germany, we have detected 17 different alleles. The most common allele was allele 13 (204 bp) with a frequency of 18.8%. Eight further alleles have frequencies higher than 5%. With a heterozygozity index of 90% and 60 different genotypes, D3S1359 has shown to be a highly polymorphic and informative marker. Sequencing data of this STR locus revealed further variation.

D3S1358: sequence analysis and gene frequency in a German population

Eight alleles of the STR system D3S1358 were observed and sequenced. The alleles ranged in size from 119 bp (13 repeats) to 147 bp (20 repeats) and consist of two diverse tetranucleotides: [AGAT] and [AGAC]. Alleles 16 and 17 show basic repeats which are different in one base. The allele frequency of 499 unrelated persons from SW-Germany yielded no deviation from Hardy-Weinberg equilibrium. A comparison with a Portuguese population showed some small differences, i.e. one allele has not been found in the Portuguese sample until now. The mutation rate was estimated with 0.8% analysing 65 families.

An architecture for distributed visual memory

The development of autonomous as well as situated robots is one of the great remaining challenges and involves a number of different scientific disciplines. In spite of recent dramatic progress, it remains worthwhile to examine natural systems, because their abilities are still out of reach. Motivated by research work done in the fields of cognitive systems, visual perception, and psychology of memory we designed and implemented a memory architecture for visual tasks. Structural and functional concepts of the memory architecture were modeled on the ones found in natural systems. We present an efficient implementation based on parallel programming techniques. The memory module is integrated into a distributed system for speech and image analysis, which is currently developed in the Sonderforschungsbereich (SFB) 360, Situated Artificial Communicators, where a hybrid vision system combining neural and semantic networks is used.

Nearby stop codons in exons of the neurofibromatosis type 1 gene are disparate splice effectors

Stop mutations are known to disrupt gene function in different ways. They both give rise to truncated polypeptides because of the premature-termination codons (PTCs) and frequently affect the metabolism of the corresponding mRNAs. The analysis of neurofibromin transcripts from different neurofibromatosis type 1 (NF1) patients revealed the skipping of exons containing PTCs. The phenomenon of exon skipping induced by nonsense mutations has been described for other disease genes, including the CFTR (cystic fibrosis transmembrance conductance regulator) gene and the fibrillin gene. We characterized several stop mutations localized within a few base pairs in exons 7 and 37 and noticed complete skipping of either exon in some cases. Because skipping of exon 7 and of exon 37 does not lead to a frameshift, PTCs are avoided in that way. Nuclear-scanning mechanisms for PTCs have been postulated to trigger the removal of the affected exons from the transcript. However, other stop mutations that we found in either NF1 exon did not lead to a skip, although they were localized within the same region. Calculations of minimum-free-energy structures of the respective regions suggest that both changes in the secondary structure of the mRNA and creation or disruption of exonic sequences relevant for the splicing process might in fact cause these different splice phenomena observed in the NF1 gene.

[The plasma level of the neurotoxin 1-trichloromethyl-1,2,4,5-tetrahydro-beta-carboline (TaClo) in man after oral administration of chloral hydrate]

Chloral hydrate (CAS 302-17-0) is a widely used hypnotic and sedative agent. It was recently reported in the literature that a neurotoxin, TaClo (1-trichloromethyl-1,2,3,4-tetrahydro-beta-carboline), may be formed in vitro from tryptamine (Ta) and chloral (Clo). Intraperitoneal administration of TaClo led to parkinson-like symptoms in the rat. Hence, the plasma levels of TaClo were determined at various time-points in 18 healthy volunteers in two periods each during a bioavailability study of several chloral hydrate preparations. The limit of quantitation for TaClo was 5 ng/ml. No TaClo could be determined in the plasma of the various volunteers following administration of human therapeutic doses of chloral hydrate. Hence, it is unlikely that TaClo will be formed in man after application of therapeutic doses of chloral hydrate to patients.

Typing of the short tandem repeat D8S347 locus with different fluorescence markers

The short tandem repeat (STR) locus D8S347 was analyzed by capillary electrophoresis. Sequencing data and a population study of 203 individuals from a southwestern German population are presented. We detected 12 different alleles, 340-388 bp in length, and found 40 different genotypes. The heterozygosity index was 85.7%. Futhermore, we investigated the consequences of different fluorescent dyes, namely 6-FAM, HEX, and ROX, on the ABI-calculated fragment sizes of defined (i.e., sequenced) alleles (348-376 bp in length). 6-FAM-labeled fragments appear to be smaller than the corresponding HEX- or ROX-labeled fragments. On average, 6-FAM-labeled fragments differ by 3.52 bp from the sequencing data, HEX-labeled ones by 2.04 bp, and ROX-labeled ones by 1.42 bp. Generally, small alleles differ less from the expected sequencing data than larger ones.

The joint development of orientation and ocular dominance: role of constraints

Correlation-based learning (CBL) has been suggested as the mechanism that underlies the development of simple-cell receptive fields in the primary visual cortex of cats, including orientation preference (OR) and ocular dominance (OD) (Linsker, 1986; Miller, Keller, & Stryker, 1989). CBL has been applied successfully to the development of OR and OD individually (Miller, Keller, & Stryker, 1989; Miller, 1994; Miyashita & Tanaka, 1991; Erwin, Obermayer, & Schulten, 1995), but the conditions for their joint development have not been studied (but see Erwin & Miller, 1995, for independent work on the same question) in contrast to competitive Hebbian models (Obermayer, Blasdel, & Schulten, 1992). In this article, we provide insight into why this has been the case: OR and OD decouple in symmetric CBL models, and a joint development of OR and OD is possible only in a parameter regime that depends on nonlinear mechanisms.

Learning and generalization in cascade network architectures

Incrementally constructed cascade architectures are a promising alternative to networks of predefined size. This paper compares the direct cascade architecture (DCA) proposed in Littmann and Ritter (1992) to the cascade-correlation approach of Fahlman and Lebiere (1990) and to related approaches and discusses the properties on the basis of various benchmark results. One important virtue of DCA is that it allows the cascading of entire subnetworks, even if these admit no error-backpropagation. Exploiting this flexibility and using LLM networks as cascaded elements, we show that the performance of the resulting network cascades can be greatly enhanced compared to the performance of a single network. Our results for the Mackey-Glass time series prediction task indicate that such deeply cascaded network architectures achieve good generalization even on small data sets, when shallow, broad architectures of comparable size suffer from overfitting. We conclude that the DCA approach offers a powerful and flexible alternative to existing schemes such as, e.g., the mixtures of experts approach, for the construction of modular systems from a wide range of subnetwork types.