Automatic ICD-10 classification of cancers from free-text death certificates

Objective Death certificates provide an invaluable source for cancer mortality statistics; however, this value can only be realised if accurate, quantitative data can be extracted from certificates – an aim hampered by both the volume and variable nature of certificates written in natural language. This paper proposes an automatic classification system for identifying cancer related causes of death from death certificates. Methods Detailed features, including terms, n-grams and SNOMED CT concepts were extracted from a collection of 447,336 death certificates. These features were used to train Support Vector Machine classifiers (one classifier for each cancer type). The classifiers were deployed in a cascaded architecture: the first level identified the presence of cancer (i.e., binary cancer/nocancer) and the second level identified the type of cancer (according to the ICD-10 classification system). A held-out test set was used to evaluate the effectiveness of the classifiers according to precision, recall and F-measure. In addition, detailed feature analysis was performed to reveal the characteristics of a successful cancer classification model. Results The system was highly effective at identifying cancer as the underlying cause of death (F-measure 0.94). The system was also effective at determining the type of cancer for common cancers (F-measure 0.7). Rare cancers, for which there was little training data, were difficult to classify accurately (F-measure 0.12). Factors influencing performance were the amount of training data and certain ambiguous cancers (e.g., those in the stomach region). The feature analysis revealed a combination of features were important for cancer type classification, with SNOMED CT concept and oncology specific morphology features proving the most valuable. Conclusion The system proposed in this study provides automatic identification and characterisation of cancers from large collections of free-text death certificates. This allows organisations such as Cancer Registries to monitor and report on cancer mortality in a timely and accurate manner. In addition, the methods and findings are generally applicable beyond cancer classification and to other sources of medical text besides death certificates.

Spectral analysis of pair-correlation bandwidth: Application to cell biology images

Images from cell biology experiments often indicate the presence of cell clustering, which can provide insight into the mechanisms driving the collective cell behaviour. Pair-correlation functions provide quantitative information about the presence, or absence, of clustering in a spatial distribution of cells. This is because the pair-correlation function describes the ratio of the abundance of pairs of cells, separated by a particular distance, relative to a randomly distributed reference population. Pair-correlation functions are often presented as a kernel density estimate where the frequency of pairs of objects are grouped using a particular bandwidth (or bin width), Δ>0. The choice of bandwidth has a dramatic impact: choosing Δ too large produces a pair-correlation function that contains insufficient information, whereas choosing Δ too small produces a pair-correlation signal dominated by fluctuations. Presently, there is little guidance available regarding how to make an objective choice of Δ. We present a new technique to choose Δ by analysing the power spectrum of the discrete Fourier transform of the pair-correlation function. Using synthetic simulation data, we confirm that our approach allows us to objectively choose Δ such that the appropriately binned pair-correlation function captures known features in uniform and clustered synthetic images. We also apply our technique to images from two different cell biology assays. The first assay corresponds to an approximately uniform distribution of cells, while the second assay involves a time series of images of a cell population which forms aggregates over time. The appropriately binned pair-correlation function allows us to make quantitative inferences about the average aggregate size, as well as quantifying how the average aggregate size changes with time.

Assessment of the underreporting rate of death cause data in Shandong province using capture-mark-recapture method [应用捕获再捕获方法评估山东省死因登记漏报率水平]

To evaluate the underreporting rate of death -cause data in Shandong province during 2012 to 2013 by capture -mark -recapture method and to provide the base for health strategy. Methods All counties were divided into 5 stratifications according the death rates of 2012, and 14 counties were selected, then 3 towns or streets were selected in each country, 10 villages or neighborhood committees were selected in each town （street）. The death data collected from security bureau and civil affairs bureau were compared with the reporting death data from the National Cause of Death Surveillance, and the underreporting rate was calculated. Results In present study, 6 929 death cases were collected, it was found that 1 556 cases were underreported. The death cases estimated by CMR method were 6 227 cases （95%CI: 7 593-7 651）, and the average underreporting rate was 23.15%. There were significantly differences between different stratifications （P<0.01）. The underreporting rate in 0-4 years old group was 56.93%, the male underreporting rate was 22.31% and the female underreporting rate was 24.09%. There was no significant difference between male and female groups （P>0.05）. Conclusion There is an obvious underreport in the cause of death surveillance of Shandong province, and the underreporting rates are different among the 5 stratifications. The underreporting rate is higher in 0-4 years old group, and the investigation of the death cause surveillance for young residents is not perfect in some countries. The investigation quality of the death cause surveillance should be improved, increasing the integrity of the report data and adjusting the mortalities in different stratifications for obtaining a accurate mortality in Shandong province.

Postcard distributed by M. Molling & Co., Hannover, to commemorate the 100th anniversary of Friedrich Schiller's death.

Addressed to Therese Gottschalk nee Molling, postmarked 25 April 1905 and mailed from Hannover to Hildesheim, most likely while Therese was visiting her Meyerhof grandparents

Zebrafish as a model of Parkinson's disease

Parkinson’s disease (PD) is the second most common neurodegenerative disease among the elderly. Its etiology is unknown and no disease-modifying drugs are available. Thus, more information concerning its pathogenesis is needed. Among other genes, mutated PTEN-induced kinase 1 (PINK1) has been linked to early-onset and sporadic PD, but its mode of action is poorly understood. Most animal models of PD are based on the use of the neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP). MPTP is metabolized to MPP+ by monoamine oxidase B (MAO B) and causes cell death of dopaminergic neurons in the substantia nigra in mammals. Zebrafish has been a widely used model organism in developmental biology, but is now emerging as a model for human diseases due to its ideal combination of properties. Zebrafish are inexpensive and easy to maintain, develop rapidly, breed in large quantities producing transparent embryos, and are readily manipulated by various methods, particularly genetic ones. In addition, zebrafish are vertebrate animals and results derived from zebrafish may be more applicable to mammals than results from invertebrate genetic models such as Drosophila melanogaster and Caenorhabditis elegans. However, the similarity cannot be taken for granted. The aim of this study was to establish and test a PD model using larval zebrafish. The developing monoaminergic neuronal systems of larval zebrafish were investigated. We identified and classified 17 catecholaminergic and 9 serotonergic neuron populations in the zebrafish brain. A 3-dimensional atlas was created to facilitate future research. Only one gene encoding MAO was found in the zebrafish genome. Zebrafish MAO showed MAO A-type substrate specificity, but non-A-non-B inhibitor specificity. Distribution of MAO in larval and adult zebrafish brains was both diffuse and distinctly cellular. Inhibition of MAO during larval development led to markedly elevated 5-hydroxytryptamine (serotonin, 5-HT) levels, which decreased the locomotion of the fish. MPTP exposure caused a transient loss of cells in specific aminergic cell populations and decreased locomotion. MPTP-induced changes could be rescued by the MAO B inhibitor deprenyl, suggesting a role for MAO in MPTP toxicity. MPP+ affected only one catecholaminergic cell population; thus, the action of MPP+ was more selective than that of MPTP. The zebrafish PINK1 gene was cloned in zebrafish, and morpholino oligonucleotides were used to suppress its expression in larval zebrafish. The functional domains and expression pattern of zebrafish PINK1 resembled those of other vertebrates, suggesting that zebrafish is a feasible model for studying PINK1. Translation inhibition resulted in cell loss of the same catecholaminergic cell populations as MPTP and MPP+. Inactivation of PINK1 sensitized larval zebrafish to subefficacious doses of MPTP, causing a decrease in locomotion and cell loss in one dopaminergic cell population. Zebrafish appears to be a feasible model for studying PD, since its aminergic systems, mode of action of MPTP, and functions of PINK1 resemble those of mammalians. However, the functions of zebrafish MAO differ from the two forms of MAO found in mammals. Future studies using zebrafish PD models should utilize the advantages specific to zebrafish, such as the ability to execute large-scale genetic or drug screens.

Reproductive Biology of Corymbia citriodora subsp. variegata and Effective Pollination across its Native Range.

The spotted gum species complex represents a group of four eucalypt hardwood taxa that have a native range that spans the east coast of Australia, with a morphological cline from Victoria to northern Queensland. Of this group, Corymbia citriodora subsp. variegata (CCV) is widespread in south-eastern Queensland and northern New South Wales. It is currently the most commonly harvested native hardwood in Queensland. However, little basic knowledge of the reproductive biology of the species is available to inform genetic improvement and resource management programmes. Here we take an integrative approach, using both field and molecular data, to identify ecological factors important to mating patterns in native populations of CCV. Field observation of pollinator visitation and flowering phenology of 20 trees showed that foraging behaviour of pollinator guilds varies depending on flowering phenology and canopy structure. A positive effect of tree mean flowering effort was found on insect visitation, while bat visitation was predicted by tree height and by the number of trees simultaneously bearing flowers. Moreover, introduced honeybees were observed frequently, performing 73% of detected flower visits. Conversely, nectar-feeding birds and mammals were observed sporadically with lorikeets and honeyeaters each contributing to 11% of visits. Fruit bats, represented solely by the grey-headed flying fox, performed less than 2% of visits. Genotyping at six microsatellite markers in 301 seeds from 17 families sampled from four of Queensland's native forests showed that CCV displays a mixed-mating system that is mostly outcrossing (tm = 0.899 ± 0.021). Preferential effective pollination from near-neighbours was detected by means of maximum-likelihood paternity analysis with up to 16% of reproduction events resulting from selfing. Forty to 48% of fertilising pollen was also carried from longer distance (>60 m). Marked differences in foraging behaviour and visitation frequency between observed pollinator guilds suggests that the observed dichotomy of effective pollen movement in spotted gums may be due to frequent visit from introduced honeybees favouring geitonogamy and sporadic visits from honeyeaters and fruit bats resulting in potential long-distance pollinations.

Characterization of the mitochondrial active-site serine protein LACTB: filaments in the mitochondrial intermembrane space

Mitochondria have evolved from endosymbiotic alpha-proteobacteria. During the endosymbiotic process early eukaryotes dumped the major component of the bacterial cell wall, the peptidoglycan layer. Peptidoglycan is synthesized and maintained by active-site serine enzymes belonging to the penicillin-binding protein and the β-lactamase superfamily. Mammals harbor a protein named LACTB that shares sequence similarity with bacterial penicillin-binding proteins and β-lactamases. Since eukaryotes lack the synthesis machinery for peptidoglycan, the physiological role of LACTB is intriguing. Recently, LACTB has been validated in vivo to be causative for obesity, suggesting that LACTB is implicated in metabolic processes. The aim of this study was to investigate the phylogeny, structure, biochemistry and cell biology of LACTB in order to elucidate its physiological function. Phylogenetic analysis revealed that LACTB has evolved from penicillin binding-proteins present in the bacterial periplasmic space. A structural model of LACTB indicates that LACTB shares characteristic features common to all penicillin-binding proteins and β-lactamases. Recombinat LACTB protein expressed in E. coli was recovered in significant quantities. Biochemical and cell biology studies showed that LACTB is a soluble protein localized in the mitochondrial intermembrane space. Further analysis showed that LACTB preprotein underwent proteolytic processing disclosing an N-terminal tetrapeptide motif also found in a set of cell death-inducing proteins. Electron microscopy structural studies revealed that LACTB can polymerize to form stable filaments with lengths ranging from twenty to several hundred nanometers. These data suggest that LACTB filaments define a distinct microdomain in the intermembrane space. A possible role of LACTB filaments is proposed in the intramitochondrial membrane organization and microcompartmentation. The implications of these findings offer novel insight into the evolution of mitochondria. Further studies of the LACTB function might provide a tool to treat mitochondria-related metabolic diseases.

Postcard distributed by M. Molling & Co. to commemorate the 100th anniversary of Friedrich Schiller's death; Hannover, Germany. Advertisements

Addressed to Therese Gottschalk nee Molling, postmarked 25 April 1905 and mailed from Hannover to Hildesheim, most likely while Therese was visiting her Meyerhof grandparents

Aenny Catzenstein nee Gottschalk shortly before her death; London.

Digital Image

Negligible evidence for regional genetic population structure for two shark species Rhizoprionodon acutus (Rüppell, 1837) and Sphyrna lewini (Griffith & Smith, 1834) with contrasting biology

Biodiversity of sharks in the tropical Indo-Pacific is high, but species-specific information to assist sustainable resource exploitation is scarce. The null hypothesis of population genetic homogeneity was tested for scalloped hammerhead shark (Sphyrna lewini, n = 237) and the milk shark (Rhizoprionodon acutus, n = 207) from northern and eastern Australia, using nuclear (S. lewini, eight microsatellite loci; R. acutus, six loci) and mitochondrial gene markers (873 base pairs of NADH dehydrogenase subunit 4). We were unable to reject genetic homogeneity for S. lewini, which was as expected based on previous studies of this species. Less expected were similar results for R. acutus, which is more benthic and less vagile than S. lewini. These features are probably driving the genetic break found between Australian and central Indonesian R. acutus (F-statistics; mtDNA, 0.751–0.903, respectively; microsatellite loci, 0.038–0.047 respectively). Our results support the spatially homogeneous monitoring and management plan for shark species in Queensland, Australia.

To reactivate or not to reactivate : Control of KSHV lytic replication is essential for apoptosis in response to p53 restoration

Kaposi's sarcoma herpesvirus (KSHV) is an oncogenic human virus and the causative agent of three human malignancies: Kaposi's sarcoma (KS), Multicentric Castleman's Disease (MCD), and primary effusion lymphoma (PEL). In tumors, KSHV establishes latent infection during which it produces no infectious particles. Latently infected cells can enter the lytic replication cycle, and upon provision of appropriate cellular signals, produce progeny virus. PEL, commonly described in patients with AIDS, represents a diffuse large-cell non-Hodgkin's lymphoma, with median survival time less than six months after diagnosis. As tumor suppressor gene TP53 mutations occur rarely in PEL, the aim of this thesis was to investigate whether non-genotoxic activation of the p53 pathway can eradicate malignant PEL cells. This thesis demonstrates that Nutlin-3, a small-molecule inhibitor of the p53-MDM2 interaction, efficiently restored p53 function in PEL cells, leading to cell cycle arrest and massive apoptosis. Furthermore, we found that KSHV infection activated DNA damage signaling, rendering the cells more sensitive to p53-dependent cell death. We also showed in vivo the therapeutic potential of p53 restoration that led to regression of subcutaneous and intraperitoneal PEL tumor xenografts without adversely affecting normal cells. Importantly, we demonstrated that in a small subset of intraperitoneal PEL tumors, spontaneous induction of viral reactivation dramatically impaired Nutlin-3-induced p53-mediated apoptosis. Accordingly, we found that elevated KSHV lytic transcripts correlated with PEL tumor burden in animals and that inhibition of viral reactivation in vitro restored cytotoxic activity of a small-molecule inhibitor of the p53-MDM2 interaction. Latency provides a unique opportunity for KSHV to escape host immune surveillance and to establish persistent infections. However, to maintain viral reservoirs and spread to other hosts, KSHV must be reactivated from latency and enter into the lytic growth phase. We showed that phosphorylation of nucleolar phosphoprotein nucleophosmin (NPM) by viral cyclin-CDK6 is critical for establishment and maintenance of the KSHV latency. In short, this study provides evidence that the switch between latent phase and lytic replication is a critical step that determines the outcome of viral infection and the pathogenesis of KSHV-induced malignancies. Our data may thus contribute to development of novel targeted therapies for intervention and treatment of KSHV-associated cancers.

New genetic loci link adipose and insulin biology to body fat distribution

Body fat distribution is a heritable trait and a well-established predictor of adverse metabolic outcomes, independent of overall adiposity. To increase our understanding of the genetic basis of body fat distribution and its molecular links to cardiometabolic traits, here we conduct genome-wide association meta-analyses of traits related to waist and hip circumferences in up to 224,459 individuals. We identify 49 loci (33 new) associated with waist-to-hip ratio adjusted for body mass index (BMI), and an additional 19 loci newly associated with related waist and hip circumference measures (P < 5 × 10−8). In total, 20 of the 49 waist-to-hip ratio adjusted for BMI loci show significant sexual dimorphism, 19 of which display a stronger effect in women. The identified loci were enriched for genes expressed in adipose tissue and for putative regulatory elements in adipocytes. Pathway analyses implicated adipogenesis, angiogenesis, transcriptional regulation and insulin resistance as processes affecting fat distribution, providing insight into potential pathophysiological mechanisms.

Genetic studies of body mass index yield new insights for obesity biology

Obesity is heritable and predisposes to many diseases. To understand the genetic basis of obesity better, here we conduct a genome-wide association study and Metabochip meta-analysis of body mass index (BMI), a measure commonly used to define obesity and assess adiposity, in up to 339,224 individuals. This analysis identifies 97 BMI-associated loci (P < 5 × 10−8), 56 of which are novel. Five loci demonstrate clear evidence of several independent association signals, and many loci have significant effects on other metabolic phenotypes. The 97 loci account for ~2.7% of BMI variation, and genome-wide estimates suggest that common variation accounts for >20% of BMI variation. Pathway analyses provide strong support for a role of the central nervous system in obesity susceptibility and implicate new genes and pathways, including those related to synaptic function, glutamate signalling, insulin secretion/action, energy metabolism, lipid biology and adipogenesis.