The NEDD family E s NEDD and Itch

The NEDD4 family E3s NEDD4 and Itch are both expressed in T-cells (Heissmeyer et al., 2004) and although in vitro data shows these ligases share a number of substrates involved in T-cell regulation, Itch and NEDD4 knockout mouse models display unique phenotypes, suggesting discrete functions in vivo ( Fang et al., 2002uanduYang et al., 2008). Foetal liver chimeras that lack NEDD4 expression in bcl-xl of haemopoietic origin show that NEDD4 is not required for T-cell development, or for initial T-cell antigen receptor mediated activation events (Yang et al., 2008). However NEDD4/ mice have fewer effector T-cells, and in response to antigen immunization T-cells lacking NEDD4 proliferate poorly, and produce less interleukin 2, suggesting the role of NEDD4 is to convert naïve T-cells into activated T-cells. The hypo-responsiveness of NEDD4/ T-cells can be explained by the impaired ubiquitination and degradation of Cbl-b, a ubiquitin ligase that plays a critical role in T-cell activation and tolerance induction, as NEDD4 is required for the poly-ubiquitination of Cbl-b ( Magnifico et al., 2003uanduYang et al., 2008). Recently it was shown that Cbl-b inhibits T-cell activation by impeding the association of NEDD4 with PTEN in T-cells to suppress PTEN inactivation (Guo et al., 2012). In addition NEDD4 is not required for B-cells to become activated, but NEDD4/ T-cells are unable to provide adequate help for B-cells to undergo immunoglobulin class switching (Yang et al., 2008).
NEDD4 is frequently overexpressed in many different types of cancer (for reviews see (Chen and Matesic, 2007uanduYe et al., 2014)). NEDD4 was first described as a proto-oncogene for its role in negatively regulating the tumour suppressor PTEN via ubiquitination in vitro ( Wang et al., 2007). An inverse correlation between (increased) NEDD4 and (decreased) PTEN has been observed in many human cancer cell lines, including breast cancer MDA-MB-231 and prostate cancer DU145 cell lines (Liu et al., 2014). After overexpressing K-ras or EGF treatment, increased NEDD4 levels and PTEN degradation are observed in various type of human cancer cell lines including cervical adenocarcinoma HeLa, colorectal adenocarcinoma HT-29, gastric adenocarcinoma BGC-823 and bcl-xl hepatocellular carcinoma HepG2 (Zeng et al., 2014). Given the lack of NEDD4 regulation of PTEN in NEDD4/ mice, this led to the hypothesis that perhaps NEDD4-mediated PTEN degradation primarily occurs in cancer cells under certain oncogenic circumstances (Zeng et al., 2014). Furthermore, the CDK-4 binding partner p34 has been identified as an interactor of NEDD4 in cancer cells lines, and co-expression of p34 and NEDD4 is correlated with lowered PTEN levels in colon cancer tissues, suggesting that NEDD4 positively regulates tumourigenesis via the p34-dependent PTEN proteasomal degradation (Hong et al., 2014). Contrary to this, there is also evidence of NEDD4 overexpression in cancer that promotes cell growth independent of PTEN signalling, such as in human colorectal cancer lines HCT-15 and LoVo (Eide et al., 2013).
Decreased levels of NEDD4 can also be associated with cancer. NEDD4 directly ubiquitinates oncoproteins N-Myc in neuroblastoma and c-Myc in pancreatic cancer cells to target these Myc proteins for proteasomal degradation (Liu et al., 2013). The histone deacetylase SIRT2 enhances expression of N-Myc and c-Myc by directly binding to the NEDD4 promoter and repressing NEDD4 gene expression by deacetylating histone H4 lysine 16 (Liu et al., 2013). Importantly, NEDD4 gene expression could be reactivated by the addition of SIRT2 inhibitors, resulting in reduced N-Myc and c-Myc protein expression, and suppressing neuroblastoma and pancreatic cancer cell proliferation (Liu et al., 2013). An inverse relationship between protein levels of (low) NEDD4 and (high) HER3 (an EGFR tyrosine kinase member) is observed in ductal cells of prostate cancer tumours compared to surrounding tissues, and knockdown of NEDD4 in human prostate and breast cancer cell lines leads to increased HER3-mediated cell migration and proliferation in vitro, and xenoplant tumour growth in vivo ( Huang et al., 2014).

In carcinoma cells the cytoskeleton is abnormal and the expression

In carcinoma cells, the cytoskeleton is abnormal and the expression of some cytoskeleton-associated proteins is aberrant (Bernal et al., 1983). In the present study, we found that SCIN is highly expressed in the lung cancer samples and about 7% lung cancer samples have aberrantly strong SCIN expression. Previous study found that T cell lysis resistant tumor cells have aberrant strong SCIN expression and knockdown SCIN could significantly attenuate the resistance to cytolytic T lymphocytes killing. Knockdown SCIN expression might facilitate lung cancer cells being killed by T lymphocytes.
In conclusion, our results suggest that SCIN silencing by lentivirus-mediated RNAi could inhibit the proliferation of lung cancer cells, which may be due to iκb/ikk inhibitor arrest and apoptosis. Our studies provide a potential therapeutic gene target for the treatment of lung cancers.
Disclosures
None.
AcknowledgmentThis work was supported by the National Natural Science Foundation of China (Grant No. 30700821), the Liaoning Bai Qian Wan Talents Program (Grant No. 2011921038), and the Liaoning Province Science and Technology (ST) Project (Grant No. 2013225585).
aa, amino acid; AO, average occurrence; CDS, coding sequence; Chr, chromosome; CWV, cold-water vibriosis; DRs, direct target repeats; IS, insertion sequence; ORF, open reading frame; Ori, replication origin; S, serine; Ter, replication terminus; Y, tyrosine
Keywords
Insertion sequence elements; Transposition; Transposase; IS microevolution lines; Target sites
1. Introduction
Outbreaks of cold-water vibriosis (CWV), which are caused by Aliivibrio salmonicida (previously Vibrio salmonicida), occur during the winteraspring period and significantly affected the salmon production industry in the past ( Lillehaug, 1990uanduSchr?der et al., iκb/ikk inhibitor 1992). In 2008, the genome of A. salmonicida was sequenced and annotated by Hjerde et al. (2008). The genome of A. salmonicida consists of two chromosomes and several plasmids (4.3ukb, 5.4ukb, 8.3ukb, 11.5ukb, 32ukb and 92ukb), which form at least 11 observed plasmid profiles (Sorum et al., 1988). Despite such plasmid diversity, no correlation between the plasmid content and the pathogenicity of the microorganism has been discovered to date ( Hjeltnes et al., 1987, Nordmo et al., 1997uanduNelson et al., 2007). Although the role of these plasmids in the lifestyle of A. salmonicida remains unclear, the functions of the chromosomes in this pathogen generally resemble those functions that have been observed in other Gammaproteobacteria ( Heidelberg et al., 2000). The majority of the genes that are located on Chromosome I (Chr_I) are involved in replication, transcription, cell division or other core functions. The smaller chromosome contains more accessory genes and, thus, is primarily involved in environmental responses and adaptability. Both chromosomes of A. salmonicida carry traces of large intra-chromosomal rearrangements when compared with other related species ( Hjerde et al., 2008). The genomic and plasmid DNA in A. salmonicida are extremely enriched by mobile genetic elements. Although phage gene sequences are relatively rare and no regions with homology to characterised transposons can be found, insertion sequence (IS) elements persist in numbers. A prediction study discovered 290 IS elements in the genome of A. salmonicida, which places it among the bacteria with the highest IS element content ( Cerveau et al., 2011). While the simultaneous transposition of IS elements quite often promotes the relocation, inversion, or excision of large DNA regions or might even lead to plasmid fusion ( Downard, 1988, Heritage and Bennett, 1985, Morita et al., 1999uanduHayes, 2003), the individual movement of IS elements might not have an impact on the host or lead to gene knockout or to an alteration of the gene expression level ( Polard et al., 1996, Mahillon and Chandler, 1998uanduTurlan et al., 2000).

Thus MEX represses translation of nos

Thus, MEX-3 represses translation of nos-2 in all going here starting at the 2-cell stage, pal-1 in the anterior at the 4-cell stage, and zif-1 in germline cells starting at the 4-cell stage. This requires that MEX-3 activity be spatially and temporally regulated for each mRNA. For example, in the P2 germline blastomere of the 4-cell embryo, MEX-3 must repress nos-2 and zif-1, but not pal-1. In anterior AB blastomeres at the same stage, MEX-3 must repress nos-2 and pal-1, but not zif-1. To achieve such a variety of repression patterns, additional factors must be involved in each case to provide specificity. For PAL-1, we know that MEX-5/6, which are localized to the anterior blastomere of the 2-cell embryo ( Schubert et al., 2000), are required along with MEX-3 for translational repression (Huang et al., 2002).
4.3. Combinatorial control
For zif-1, translational regulation in the embryo requires MEX-3, SPN-4, MEX-5/6, and POS-1 ( Oldenbroek et al., 2012). The available evidence suggests that MEX-3 and SPN-4 inhibit translation of zif-1 in all cells of the early embryo, and then POS-1 takes over repression in the germline blastomeres ( Oldenbroek et al., 2012). In somatic cells, MEX-5/6 relieve repression by competing with POS-1 and possibly MEX-3 for binding to the zif-1 3UTR ( Oldenbroek et al., 2012).
While regulation of pal-1, nos-2, and zif-1 in the embryo all require MEX-3, SPN-4, and MEX-5/6, there are also some notable differences. PIE-1 only affects nos-2 expression. POS-1 only has a clear effect on nos-2 and zif-1 expression. [Depletion of pos-1 results in variable abnormal PAL-1 expression, which suggests that any effects are indirect ( Huang et al., 2002)]. The particular combination of factors present on a 3UTR may control the activity of these factors, for example causing POS-1 in the germline to activate nos-2 expression but repress zif-1 expression, and ultimately produce unique patterns of temporal and spatial gene expression.
Modification of a regulatory factor may change its effect on one target but not another. For example, par-4 dependent inactivation of MEX-3 in the posterior at the 4-cell stage prevents it from repressing pal-1, but not nos-2 (which is repressed in all cells at that stage). Thus, translational regulation in the early embryo seems to rely on a relatively small number of RBPs, which can produce a variety of temporal and spatial patterns of gene expression by acting in different combinations, and changing their activity over time.
In addition to its role in embryogenesis, several lines of evidence implicate C. elegans MEX-3 in maintenance of the germline stem cells. First, MEX-3 is present in mitotic germline stem cells ( Ciosk et al., 2004), where it is required to repress translation of at least one transcript characteristic of differentiating oocytes (Ciosk et al., 2006). In addition, MEX-3 functions redundantly with PUF-8 to promote mitosis of these germline stem cells (Ariz et al., 2009). Furthermore, MEX-3 and GLD-1 are together required to prevent transdifferentiation of developing oocytes into somatic cell types, thereby maintaining germline totipotency (Ciosk et al., 2006). Notably, most of the ectopic body muscle produced from the mex-3() gld-1() germline requires pal-1 activity, indicating that transdifferentiation is due to premature activation of maternally provided cell fate determinants ( Ciosk et al., 2006).

RNA isolation and Illumina sequencing Total RNA

2.2. RNA isolation and Illumina sequencing
Total RNA was isolated from a whole single adult of E. pancreaticum using TRIzol reagent (Invitrogen, Life Technologies, Carlsbad, CA, USA) according to the manufacturer\’s protocol. Total RNA of independent adult E. pancreaticum was stored at u80uC until use. The Oligo (dT) was used to isolate poly (A) mRNA from total RNA. Mixed with the fragmentation buffer, the mRNA was fragmented into short fragments. Then cDNA was synthesized using the mRNA fragments as templates. Short fragments were purified and resolved with EB buffer for end reparation and single nucleotide A (adenine) addition. After that, the short fragments were connected with adapters. The suitable fragments were selected for the AS1404 amplification as templates. During the QC steps, Agilent 2100 Bioanaylzer and ABI StepOnePlus Real-Time PCR System were used in quantification and qualification of the sample library. Illumina HiSeqTM 2000 was used for sequencing at the BGI-Shenzhen, Shenzhen, China according to the manufacturer\’s instructions (Illumina, San Diego, CA, USA).
Prior to assembly, the high-quality clean reads were obtained from raw reads by removing adaptor sequences, highly redundant sequences, reads that contained more than 10% N; rate (the N; character representing ambiguous bases in reads), and low quality reads containing more than 50% bases with Q-valueuu20. De novo assembly of the clean reads was performed by using the Trinity software, which was designed specifically for transcriptome assembly ( Grabherr et al. 2011). Briefly, Trinity first combines reads of a certain length of overlap to form longer fragments, which are called contigs. Then the reads are mapped back to contigs. Trinity connects the contigs and gets sequences that cannot be extended on either end. Such sequences are defined as unigenes. Unigenes from each sample\’s assembly can be used in further processes of sequence splicing and redundancy removal with sequence clustering software TGICL (Pertea et al. 2003) in order to acquire non-redundant unigenes that are as long as possible.
2.4. Bioinformatics analysis
Unigene sequences were first aligned to the protein databases NT, NR, Swiss-Prot, Cluster of Orthologous Groups (COG) and KEGG databases by BLASTx, using an e-valueu<u0.00001. The unigenes were tentatively annotated according to the known sequences with the highest sequence similarity. The annotated unigenes direction and CDSs were identified by the best alignment results. ESTScan (Iseli et al. 1999) was used to predict the coding sequences (CDS) and the sequence direction when unigenes were unaligned to any of the databases. With nr annotation, the Blast2GO program was used to classify unigenes to GO terms such as molecular function, biological processes, and cellular components (Conesa et al. 2005). After obtaining GO annotations for all unigenes, WEGO software (Ye et al. 2006) was used to perform GO function classification for all unigenes and to analyze the distribution of E. pancreaticum gene functions AS1404 at the macro level. Simple sequence repeats (SSRs) in the nucleotide sequences were identified using the MIcroSAtellite identification tool (MISA) ( Thiel et al. 2003). The poly-A and poly-T sequences at the terminal regions of the UTs were removed before SSR identification. SOAPsnp (Li et al. 2009) was used (with parameters -u t -Q i -L 90) on pileup files to output lists of single nucleotide polymorphisms (SNPs) and their locations.
3. Results
3.1. Sequencing and assembly
Fig. 1.uLength distribution of unigenes of adult Eurytrema pancreaticum transcriptome.Figure optionsDownload full-size imageDownload high-quality image (180 K)Download as PowerPoint slide
3.2. Functional annotation
A total of 5510 (23.4%) unigenes were assigned 5,555 GO term annotations, which could be classified into three categories: biological process, molecular function, and cellular component. The biological process category consisted of 3,624 GO terms, which were assigned to 4,204 (17.8%) unigenes. The cellular component category consisted of 707 GO terms, which were assigned to 4,227 (17.9%) unigenes, and the molecular function category consisted of 1,224 GO terms, which were assigned to 3,359 (14.2%) unigenes (Fig. 2). Within the biological process category, most unigenes were assigned to “cellular process” (963 terms). In the cellular component category, most unigenes were assigned to “cell” (285 terms). In the molecular function category, the major GO terms were “binding” (408 terms) (Fig. 2).

Some miRNAs whose expressions are not limited to the testis

Some miRNAs, whose expressions are not limited to the testis, may also show some involvement in the antenatal or postpartum development of male germ cells. One example is miR-103 which has various roles in different tissues but also positively regulates pre-adipocyte differentiation in the infant porcine testes (Li et al., 2011). Some mouse testis-specific miRNAs are markedly expressed in Sertoli 6XHis and are under the regulation of androgen (Panneerdoss et al., 2012). The active performance of miRNAs in spermatogenesis are not restricted to mammals. In the chicken, for example, miR-202 shows importance for sexual differentiation in male gonads (Bannister et al., 2009).
2.3. miRNA related proteins in spermatogenesis
Since miRNAs undergo multiple aspects of processing and are not the unique components of miRISC, protein factors in the miRNA pathway are also probable contributors to male germ cell differentiation and development. Unlike testis-specific miRNAs, which directly exert post-transcriptional regulation roles significant to spermatogenesis, the proteins contributing to miRNA biogenesis and regulation abilities can also help indirectly disclose the association between miRNAs and spermatogenesis.
Among those proteins, Dicer is the most studied due to its central role in miRNA maturation. Defects in Dicer in the male germline will lead to a decrease of germ cells and the inhibition of maturation by preventing the ongoing development of round spermatids into elongating spermatids, and finally causing asthenospermia and infertility (Maatouk et al., 2008). This is because Dicer participates in chromatin structural organization and nuclear shaping during the round to elongated spermatid stage transition (Korhonen et al., 2011). Deletion of Dicer in postnatal mouse testis also decreases the expression of sex chromosome encoded genes and causes a large proportion of germ cells to arrest at the prepachytene stage (Greenlee et al., 2012). Even a Dicer-null mutation in PGCs and spermatogonia will also lead to proliferation inhibition ( Hayashi et al., 2008). Dicer also participates in chromatin remodeling in haploid germ cells and in the elongation process of spermatids (Wu et al., 2012). Dicer is also indispensible to maintain a normal seminiferous epithelium structure. The deletion of the Dicer gene in Sertoli cells also results in infertility in the male mouse ( Kim et al., 2010uanduPapaioannou et al., 2009). It is also clear that miRNAs influence a large spectrum of Sertoli cellular proteins that have significant supporting roles in male reproduction (Papaioannou et al., 2011). However, since Dicer is shared by both miRNA and siRNA pathways, there is a possibility that endogenous small-interfering RNAs (endo-siRNAs) are also involved in spermatogenesis regulation. This will be discussed in the fourth part of this review. Mice expressing defective DGCR8 also show infertility phenotypes, although the defects are less severe than that of Dicer deficiency (Zimmermann et al., 2014). In addition to Dicer and DGCR8, Drosha and AGO4 are also important to spermatogenesis. Drosha is crucial to the formation of normal spermatozoa. Its deficiency can cause abnormalities in both meiotic and post-meiotic germ cells (Wu et al., 2012). AGO4, enriched in the nucleus of spermatocytes, guards the gateway from mitosis into meiosis. The loss of its function reduces the expression of some miRNA species in meiotic prophase I germ cells (Modzelewski et al., 2012).

Some origins of morphological misinterpretation can be traced

Some origins of morphological misinterpretation can be traced back, at least hypothetically, to series of sequential illustrations that were based on each other. For example, the ears of the elephant are often portrayed as fan-shaped, ribbed structures. This could be due to the fact that the ears of the elephants, especially those of the African elephant (L. africana), are permeated with large blood vessels. These swell to cool down excess body heat and then form a structure similar to a net or fan. If the first illustrations of this type were based on an authentic representation of these vein systems, it could be concluded that this detail was expanded in later illustrations, until it had finally developed into a distinctive fan shape.
A similar process could also be the cause of the outwardly open concave elephant ears, which are mainly found in the naturalistic illustrations of the Renaissance. The ear canal of the elephant actually does lie on the outside of the ear, however the ear tends to lay flat and does not have a concave shape like human or horse ears do. Though, due to the shape of the elephants head there is a shadow enough on its ear, which was perhaps interpreted as an opening. From copy to copy this feature has become more pronounced until the elephant ear had developed into a very pronounced concave shape.
3.2. Parallels to biological organisms and to evolutionary processes
Some of the illustrated elephants possess a unique morphology that does not repeat itself, at least not within my collection. However, most pictures have similarities, suggesting a common origin or a family relationship. Until the invention of printing technologies such as wood print or copper engraving, books and the illustrations contained within them were duplicated laboriously by hand. This process led to variations, not exact copies. So also the morphology of the elephant changed continually with each reproduction and formed new features, while other features were lost in this process. Certainly different images also merged into each other and formed hybrids that combined the features of several earlier pictures. The illustrated elephants behave in this process enough quite like biological organisms and species, in the sense that they have an individual anatomy, they crossbreed, propagate, evolve and develop new traits and characteristics. For this reason, it is possible to describe the Elephas anthropogenus according to taxonomic principles, and to reinterpret its development from the perspective of evolutionary biology.
3.3. The classification of the Elephas anthropogenus
For the integration of the elephant depictions into a pseudo-biological system, the images in my collection went through several phases and transformations. First I positioned the images in chronological order, through which its development along a time line became visible.
As a next step, I made simplified line drawings from the elephant illustrations (Fig. 1). The transfer of the illustrations, which were originally crafted in very different styles, into an uniform, reduced representation was necessary to allow for a clear visual comparison, which focused on the actual morphology of the elephant. The isolation of individual species from their surrounding context is also a fundamental method of the natural sciences.

If we use explicitly all the scenarios induced by

If we use explicitly all the scenarios induced by the problem, we obtain a model with a number of integer variables which, in the worst case, is equal to 2 J + I × J × Ω 2 J + I × J × Ω . In practice, J J is typically not too large. However this is not the case with I I and, clearly not with Ω Ω , as we noted before. In the next section we use these figures together with an illustrative example to show that the above model is intractable for (large) real-world instances. Nevertheless, by understanding the shortcomings of the model we also propose a way for using it contact us and devise an approximate approach for the problem.
As a final comment to the model , , , , , , and we would like to point out that it is more general that it seems at a first glance. First, the costs fjfj and gj(j∈J) may include the operation or maintenance of the bases and ambulances, respectively. Second, we may not wish to build a system from scratch as we have assumed so far. In this case, some bases are already operating with some ambulances allocated to them. Such a situation can be easily accommodated in the modeling framework proposed by setting to 1 the location variables associated with those bases. Furthermore, the zz-variables for those bases will now represent the number of new ambulances to allocate to the base. Accordingly, constraints and should be adjusted by summing to zjzj the number of ambulances that is already operating in the corresponding base. Regarding the costs, we can simply consider the operating costs for the existing bases and add them as a constant term to the objective function. We can do the same with the operating costs for the existing ambulances.
3. Illustrative example
Consider one instance represented by the graph depicted in Fig. 2. In this case, we have 6 demand points. We assume that they coincide with the potential locations for the bases (I=JI=J).
Fig. 2. Instance with 4 demand nodes.Figure optionsDownload full-size imageDownload as PowerPoint slide
An edge between two nodes indicates that each node can cover the other regarding the maximum allowed response time. Additionally, if an emergency occurs, each node requires or 1 ambulances with the probabilities depicted in the figure. As stated in the previous section, we assume that demands are independent.
In order to put the emphasis on the relevant aspects associated with the scenarios, we assume that all costs for ambulances and bases are equal to 1. After identifying all the scenarios as well as the corresponding probabilities, model , , , , , , and was loaded and solved using the IBM ILOG CPLEX Optimization Studio. In the optimal solution bases 3 and 5 are selected with 1 and 3 ambulances located there, respectively.
The size of this instance allows us to solve it to optimally using an off-the-shelf solver. However, if the dimensions in terms of I I or Ω Ω increase significantly, then we may have to resort to some simplification and thus to approximating the optimal solution. This possibility was contact us considered for this small instance. The most obvious simplification consists of working with one sample of scenarios instead of working with the entire set. The dimension of a sample ranges between 1 and Ω Ω .

by using the phrase ldquo Tyche the

by using the phrase “Tyche, the goddess of chance, picks a sample” to describe this choice (see for instance [2] and [60]).2.Symbolic ↔ probability space.
Conceptually, the σ -algebra FF of a probability space contains the universe of all the yes/no questions (i.e. propositions) that the subject can entertain. A particular aspect of a given state of Nature ω is extracted via a corresponding random variable X:Ω→X,X:Ω→X, mapping ω into a symbol X(ω ) from a set of symbols XX. Random variables can be combined to form complex aspects, and the ensuing symbols are consistent (i.e. free of contradictions) as guaranteed by construction. Thus, a probability space and the associated collection of random variables make up the structure of the potential realities that the subject can hope to comprehend. Furthermore, one can associate to each random variable at least one of three rôles (but typically just one), detailed next.3.Imaginary ↔ hypotheses.
A random variable can play the rôle of a latent feature of the state of Nature. Latent variables furnish the sensorimotor space with a conceptual or signifying structure, and a particular configuration of these variables constitutes a alk inhibitor in the Bayesian sense. Because of this function, we can associate the collection of latent variables to Lacan’;s imaginary register.4.Flow between I and the Other ↔ actions & observations.
The hypotheses by themselves do not ground the subject’;s symbolic domain to any reality however—for this, variables modelling interactions are required. These variables capture symbols that appear in the sensorimotor stream of the subject, that is, at her boundary with the world, modelling the directed symbolic flow occurring between the I and the Other; in particular, the out- and inward flows are represented by actions and observations, respectively.5.Objet petit a ↔ causal intervention.
The last connection I would like to establish, which will become a central theme in what follows, is between the object petit a and causal interventions. Lacanian theory explains agency in terms of a kink in the signifying chain—that is, the interruption of a pre-existing relation between two symbols—that is subjectivised in hindsight [14] and [58]. One crucial aspect of this notion is that it requires the comparison between two instants of the signifying network, namely the one where the relation is still intact and the resulting one where the relation is absent, adding a dynamic element to the static symbolic order. This element has no analogue in standard probability theory. However, the last twenty years have witnessed the systematic study of what appears to be an analogous idea in the context of probabilistic causality. More precisely, the interruption of the signifying chain is a causal intervention [38].

Table SM in the supplementary material shows the

Table SM7 in the supplementary material shows the emission factor of 2,3,7,8-Br substituted PBDD/Fs and total WHO2005-TEQ for the emissions of the different combustion and pyrolysis runs performed from the ASR. The most abundant isomer was 1,2,3,4,6,7,8-HpBDF either in combustion and pyrolysis at 600 °C as also reported by Ortuño et al. treating electronic wastes (Ortuño et al., 2014).
The run with the maximum total emission of PBDD/Fs was that of combustion at 600 °C with a λ = 0.3. This is consistent with the study reported by Conesa et al. (2009) in which a comparison between the emission rates from pyrolysis and combustion of different wastes was presented, showing that the presence of a small amount of oxygen could promote the formation of some pollutants.
No data were found in literature in which brominated dioxin formation in ASR thermal treatments was reported so it was not possible to do a comparison.
4. Conclusions
In order to better understand the environmental impact during the pyrolysis and combustion of ASR, runs from pure pyrolysis to over-stoichiometric combustion at two different temperatures in a laboratory scale reactor were performed. The products have been analyzed and quantified in order to evaluate whether the thermal valorization of ASR might be feasible. The main conclusions that we can extract from the study are:?Significant amounts of CO and CO2 are generated in pyrolysis. These gases come from the decomposition of polymers (polycarbonates and polyurethanes) present in the ASR and from the decomposition of carbonate fillers present in automotive plastics.?The combination of high temperatures and high oxygen ratio causes the drastic inorganic ions of light hydrocarbons.?The maximum emissions of the 16 priority PAHs are obtained in pyrolysis at 850 °C. The most abundant compound, in all experiments, is naphthalene.?Chlorobenzenes, chlorophenols and bromophenols yields are almost negligible.?The greatest yields of PCDD/Fs, dl-PCBs and PBDD/Fs are obtained at 600 °C. Furans contribute more than dioxins to total emissions.
Based on the results obtained in this study, it can be stated that thermal recovery may be a feasible method for accomplishing the recovery rates established in the Directive 2000/53/EC.
AcknowledgmentsSupport for this work was provided by the CTQ2013-41006-R project from the Ministry of Economy and Competitiveness (Spain) and the PROMETEOII/2014/007 project from the Valencian Community Government (Spain). The authors are grateful to CEMEX ESPAÑA, S.A. for supplying the samples.
End-of-life vehicles; Manual dismantling; Fuzzy analytic hierarchy process; Dismantling scenario
1. Introduction
The number of automobile owners worldwide has reached 1 billion, and the number of end-of-life vehicles (ELVs) is estimated to be approximately 60 million. Such a large number elicits people’;s attention because global resources are decreasing (Chen and Zhang, 2009 and Go et al., 2011). Many countries face the challenge of rationally and efficiently utilizing these resources (Coates and Rahimifard, 2008). Various countries have developed appropriate policies and regulations to handle ELVs. Directive 2000/53/EC of the European Parliament and the European Council (September 18, 2000) is considered the first global policy for ELVs. This directive clearly regulates the responsibility of car manufacturers for recycling ELVs (Gerrard and Kandlikar, 2007). Directive 2000/53/EC required that no later than 1 January 2015, for all end-of life vehicles, the reuse and recovery rate shall be increased to a minimum of 95% by an average weight per vehicle and year. Within the same time limit, the re-use and recycling rate shall be increased to a minimum of 85% by an average weight per vehicle and year. Unlike the EU countries, Japan, and Korea, the end-of-life vehicle recovery in United States is driven by the market rather than by government regulation. The ELV management activities have been mainly impacted from national legislations (like Clean Air Act) addressing solid and hazardous waste disposals such as banning the disposal of free liquids in landfills and banning the disposal of lead acid batteries in landfills rather than a specific ELVs directive (Amelia et al., 2009).

The method we propose is different First we

The method we propose is different. First, we propose a new uncertainty measure that improves the performance of the active learner compared to the commonly used uncertainty measures. The proposed strategy determines the smallest importance weight required for the prediction to switch to another label. If a small weight is sufficient to change the prediction, then the classifier is uncertain, and the true label is queried from a labeller. Second, we propose an adaptive uncertainty threshold which is convenient for evolving streams and gives a compromise between the error rate and the number of queried labels.
This paper is organised as follows. In Section 2 we give background on active learning by focusing on the uncertainty based strategies. In Section 3 we first describe our proposed uncertainty query strategy based on instance weighting, then, we propose an adaptive threshold for a stream-based active learning. In Section 4 we present the experimental evaluation. Finally, we conclude and explain the future work in Section 5.
2. Preliminaries
Let X?RdX?Rd be a d dimensional feature space. The input x ∈ X is called an instance and represents a data point in the feature space X. Let Y be a finite set of Tenovin-6 where each class y ∈ Y is presented as a discrete value called class label. The classifier is then presented as a classification function h that associates an instance x ∈ X to a class y ∈ Y (see Eq. (1)).
equation(1)h: X?Yx?y=h(x)
The conditional probability p(y x) is called the posterior probability of class y for instance x. A discussion of specific classifiers properties is beyond the scope of this paper. We simply mention that most existing classifiers not only return the predicted class y but also gives a score or an estimate p^(y x) of the posterior probability. The main question that arises is how to get a good classifier h, i.e., how to learn h.
In the case of a supervised learning, the algorithm tries to model the relationship between inputs (instances) and outputs (classes) by learning the function h:X?Yh:X?Y using a training dataset where each instance is previously labelled with its true class label. Suppose that we have a large amount of unlabelled instance. A labeller will manually label as much instances as possible to use them as a training dataset. The more we label, the better adaptation is. However, labelling is costly and time consuming. Note that the instances to be labelled are randomly selected by the labeller, i.e., the learning is passive.