Identification of native American founder mtDNAs through the analysis of complete mtDNA sequences: Some caveats

In this study, a detailed analysis of both previously published and new data was performed to determine whether complete, or almost complete, mtDNA sequences can resolve the long-debated issue of which Asian mtDNAs were founder sequences for the Native American mtDNA pool. Unfortunately, we now know that coding region data and their analysis are not without problems. To obtain and report reasonably correct sequences does not seem to be a trivial task, and to discriminate between Asian-and Native American mtDNA ancestries may be more complex than previously believed. It is essential to take into account the effects of mutational hot spots in both the control and coding regions, so that the number of apparent Native American mtDNA founder sequences is not erroneously inflated. As we report here, a careful analysis of all available data indicates that there is very little evidence that more than five founder mtDNA sequences entered Beringia before the Last Glacial Maximum and left their traces in the current Native American mtDNA pool.

Genetic diversity and relationship of Yunnan native cattle breeds and introduced beef cattle breeds

In this study, random amplified polymorphic DNA (RAPD) analysis was used to estimate genetic diversity and relationship in 134 samples belonging to two native cattle breeds from the Yunnan province of China (DeHong cattle and DiQing cattle) and four intro

Prion protein gene (PRNP) polymorphisms in native Chinese cattle

Le polymorphisme au sein de quatre regions du gene codant pour la proteine prion bovine (PRNP) confere la susceptibilite a l'encephalopathie bovine spongiforme (BSE). Ceux-ci comprennent un polymorphisme d'insertion/deletion (indel) de 23 pb dans le promoteur, un indel de 12 pb dans l'intron 1, un octapeptide repete ou un indel de 24 pb au sein du cadre de lecture, et un polymorphisme mononucleotidique (SNP) dans la region codante. Dans ce travail, les auteurs ont examine la frequence des genotypes, des alleles et des haplotypes pour ces indel au sein de 349 bovins d'origine chinoise, de meme que la sequence nucleotidique de ce gene chez 50 de ces animaux. Leurs resultats montrent que l'allele ayant la deletion de 12 pb et l'haplotype combinant la deletion de 23 pb et la deletion de 12 pb, lesquels ont ete suggeres comme etant importants pour la susceptibilite a la BSE, sont rares au sein des bovins du sud de la Chine. Une difference significative a ete observee entre les bovins affectes par la BSE et les bovins chinois sains pour ce qui est de l'indel de 12 pb. Au total, 14 SNP ont ete observes dans la region codante du gene PRNP chez les bovins chinois. Trois de ces SNP etaient associes a des changements d'acides amines (K3T, P54S et S154N). La substitution E211K qui a ete rapportee recemment chez un cas atypique de la BSE chez un bovin americain n'a pas ete detectee dans ce travail.

Distribution of endangered native fish fauna and their conservation in Lake Victoria

Lake Victoria in East Africa, supports socio-economically important fisheries for more than 30 million inhabitants in the lake basin. The lake had until the 1970's a diverse fish assemblage dominated by haplochromines species which formed at least 83% of the fish biomass (Kudhongania & Cordone 1974). The more than 500 haplochromine species in Lake Victoria, over 99% of them endemic, exploited virtually all the food sources in the lake (Witte and van Oijen 1990). Each species had its own unique combination of food and habitat preference (Goldschmidt et al., 1990).

The biology, ecology, distribution and conservation of some surviving native non-cichlid fish species of Lakes Victoria, Kyoga and Nabugabo

Prior to introduction of non-native fish species into Lakes Victor i a, Kyoga and Nabugabo, the three lakes suppor ted diverse fish fauna representing 13 families consisting of six cichlid genera and fifteen non-cichlid genera. There were about 50 non-cichlid species and over 300 cichlids consisting of mainly haplochromines (Graham 1929, worthington 1929, Greenwood 1960). Many of the species were commercially and scientifically important and provided a rich variety of protein source to choose from. Following introduction of the Nile perch and several tilapiines species, most of the native species were drastically reduced and some have apparently disappeared. The few remaining species appear to be restricted in distribution due to the presence of the Nile perch. They are mainly confined to refugia such as marginal macrophytes, rocky outcrops and small satellite lakes which are separated from the areas of introduction by swamps

The biology and ecology of native non-cichlids in the Victoria and Kyoga lake basins

An overview of the biology and ecology of some of the constantly less important commercial species is given below. These included Bagrus docmac, Clarias gariepinus, Protopterus aethiopicus, Labeo victorianus, Barbus spp, Mormyrids, Synodontis spp, and Schilbe intermedius. The stocks of most of these species declined due to over-exploitation and introduction of non-native fishes especially Nile perch. A few of these taxa still survive in the main lake and others in satellite lakes. The current status of these species in the Victoria lake basin is not known but the available information provided some information on some habitat and other requirements of some of these originally important species of the Victoria lake basin.

THE ORIGIN AND GENETIC DIFFERENTIATION OF NATIVE BREEDS OF PIGS IN SOUTHWEST CHINA - AN APPROACH FROM MITOCHONDRIAL-DNA POLYMORPHISM

Mitochondrial DNA (mtDNA) of six breeds of native domestic pigs from Yunnan province, southwest China, and two wild boars obtained from Sichuan, China, and Vietnam was analyzed using 20 restriction endonucleases that recognize six nucleotides. Restriction maps were made by double-digestion methods and polymorphic sites were located on the map. According to their mtDNA restriction types, all the breeds were classified into six groups. Genetic distances among groups were calculated to define their phylogenetic relationships. The relationship between the Sichuan wild boar and domestic pigs is close, while the Vietnamese wild boar is relatively far from them, so the domestic pigs in southwest China are likely to have originated from a wild pig which distributed in west China. We compare our results with previous reports in literature and discuss the relationship among Chinese pigs, Japanese pigs, and European pigs. The mtDNA cleavage pattern of the Mingguang pig digested by EcoRV was identical to that of Duroc; mutations at the EcoRI site, detected in the mtDNA of two Dahe pigs, are the same as in the Vietnamese wild boar, suggesting that mutational hot spots exist in the mtDNA of pigs.

Statistical parametric speech synthesis based on speaker and language factorization

An increasingly common scenario in building speech synthesis and recognition systems is training on inhomogeneous data. This paper proposes a new framework for estimating hidden Markov models on data containing both multiple speakers and multiple languages. The proposed framework, speaker and language factorization, attempts to factorize speaker-/language-specific characteristics in the data and then model them using separate transforms. Language-specific factors in the data are represented by transforms based on cluster mean interpolation with cluster-dependent decision trees. Acoustic variations caused by speaker characteristics are handled by transforms based on constrained maximum-likelihood linear regression. Experimental results on statistical parametric speech synthesis show that the proposed framework enables data from multiple speakers in different languages to be used to: train a synthesis system; synthesize speech in a language using speaker characteristics estimated in a different language; and adapt to a new language. © 2012 IEEE.

Phrase-based statistical language generation using graphical models and active learning

Most previous work on trainable language generation has focused on two paradigms: (a) using a statistical model to rank a set of generated utterances, or (b) using statistics to inform the generation decision process. Both approaches rely on the existence of a handcrafted generator, which limits their scalability to new domains. This paper presents BAGEL, a statistical language generator which uses dynamic Bayesian networks to learn from semantically-aligned data produced by 42 untrained annotators. A human evaluation shows that BAGEL can generate natural and informative utterances from unseen inputs in the information presentation domain. Additionally, generation performance on sparse datasets is improved significantly by using certainty-based active learning, yielding ratings close to the human gold standard with a fraction of the data. © 2010 Association for Computational Linguistics.

Improving LVCSR system combination using neural network language model cross adaptation

State-of-the-art large vocabulary continuous speech recognition (LVCSR) systems often combine outputs from multiple subsystems developed at different sites. Cross system adaptation can be used as an alternative to direct hypothesis level combination schemes such as ROVER. The standard approach involves only cross adapting acoustic models. To fully exploit the complimentary features among sub-systems, language model (LM) cross adaptation techniques can be used. Previous research on multi-level n-gram LM cross adaptation is extended to further include the cross adaptation of neural network LMs in this paper. Using this improved LM cross adaptation framework, significant error rate gains of 4.0%-7.1% relative were obtained over acoustic model only cross adaptation when combining a range of Chinese LVCSR sub-systems used in the 2010 and 2011 DARPA GALE evaluations. Copyright © 2011 ISCA.

Use of contexts in language model interpolation and adaptation

Language models (LMs) are often constructed by building multiple individual component models that are combined using context independent interpolation weights. By tuning these weights, using either perplexity or discriminative approaches, it is possible to adapt LMs to a particular task. This paper investigates the use of context dependent weighting in both interpolation and test-time adaptation of language models. Depending on the previous word contexts, a discrete history weighting function is used to adjust the contribution from each component model. As this dramatically increases the number of parameters to estimate, robust weight estimation schemes are required. Several approaches are described in this paper. The first approach is based on MAP estimation where interpolation weights of lower order contexts are used as smoothing priors. The second approach uses training data to ensure robust estimation of LM interpolation weights. This can also serve as a smoothing prior for MAP adaptation. A normalized perplexity metric is proposed to handle the bias of the standard perplexity criterion to corpus size. A range of schemes to combine weight information obtained from training data and test data hypotheses are also proposed to improve robustness during context dependent LM adaptation. In addition, a minimum Bayes' risk (MBR) based discriminative training scheme is also proposed. An efficient weighted finite state transducer (WFST) decoding algorithm for context dependent interpolation is also presented. The proposed technique was evaluated using a state-of-the-art Mandarin Chinese broadcast speech transcription task. Character error rate (CER) reductions up to 7.3 relative were obtained as well as consistent perplexity improvements. © 2012 Elsevier Ltd. All rights reserved.

Syllable language models for Mandarin speech recognition: exploiting character language models.

Mandarin Chinese is based on characters which are syllabic in nature and morphological in meaning. All spoken languages have syllabiotactic rules which govern the construction of syllables and their allowed sequences. These constraints are not as restrictive as those learned from word sequences, but they can provide additional useful linguistic information. Hence, it is possible to improve speech recognition performance by appropriately combining these two types of constraints. For the Chinese language considered in this paper, character level language models (LMs) can be used as a first level approximation to allowed syllable sequences. To test this idea, word and character level n-gram LMs were trained on 2.8 billion words (equivalent to 4.3 billion characters) of texts from a wide collection of text sources. Both hypothesis and model based combination techniques were investigated to combine word and character level LMs. Significant character error rate reductions up to 7.3% relative were obtained on a state-of-the-art Mandarin Chinese broadcast audio recognition task using an adapted history dependent multi-level LM that performs a log-linearly combination of character and word level LMs. This supports the hypothesis that character or syllable sequence models are useful for improving Mandarin speech recognition performance.