21kPa. Diabetic ketoacidosis was diagnosed and treated

according to hospital guidelines. Over the next six hours, the patient’s symptoms rapidly improved. At the time of diagnosis of DKA, cardiotocograph (CTG) monitoring was pathological with reduced baseline variability which returned to normal within 24 hours of initiation of DKA treatment. Following treatment of DKA, bicarbonate level rose to 17mmol/L and remained at this find more level until delivery, two weeks later. The patient went into spontaneous labour; however, in view of the suspected macrosomia, she underwent an uncomplicated lower segment caesarean section resulting in the delivery of a live female weighing 4.34kg with APGAR scores of 8, 9 and 10 at 1, 5 and 10 minutes, although the cord pH was 6.9. The baby had severe neonatal hypoglycaemia, with blood glucose 1.5mmol/L, necessitating admission to the neonatal intensive care unit and treatment with intravenous dextrose for 48 hours. The patient had a six-week postpartum OGTT which showed impaired glucose tolerance with a fasting glucose of 3.9mmol/L and a two-hour glucose of 9.3mmol/L. Both mother and child were well at last contact. This case highlights the fact that SGI-1776 molecular weight women with GDM are at risk of developing DKA in later pregnancy. Recognised risk factors for DKA in pregnant patients with T1DM include infection, vomiting, treatment non-compliance, new onset diabetes, insulin pump failure,

corticosteroids and beta-adrenergic drugs.1,3 The likely precipitant in this case of GDM was administration of corticosteroids. The use of steroids in patients Dehydratase with DM is associated with a significant worsening of glycaemic

control for up to 48 hours after steroid administration.6 In pregnant women with DM who receive antenatal steroids, blood glucose control can be achieved with additional insulin, either calculated according to prior insulin requirements or via an insulin sliding scale.6,7 Venous glucose at diagnosis of DKA may be considerably lower in pregnant than in non-pregnant women. In one case-control study the blood glucose levels were compared in 90 pregnant and 286 non-pregnant females at diagnosis of DKA, and were found to be significantly lower in pregnant women with DKA: 16.3±4.6 and 27.5±4.8mmol/L (mean±SD), respectively.8 The reduced glucose level at presentation of DKA in pregnant women with DM, as in this case, may present diagnostic difficulties. All patients with DM, including those with GDM, who are unwell or present with any combination of nausea, vomiting or reduced calorific intake, should be assessed for the possibility of DKA9 with serum urea and electrolytes, venous blood gases and testing of blood or urine for ketones regardless of blood glucose readings. Acid-base balance in pregnancy is characterised by a physiological hyperventilatory response leading to a primary respiratory alkalosis.

, 2000). In this study, deletion of the orthologous gene Mga1 in fermentation fungus M. ruber M7 enhanced both citrinin and pigment production. Although the role of Mga1 in regulating mycotoxin in M. ruber M7 is consistent

with that in Aspergillus spp., the regulation role in pigment production is different from cpg-1 in C. parasitica, as disruption of cpg-1 leads to significant reductions in pigmentation (Gao & Nuss, 1996; Hicks et al., 1997; Tag et al., 2000). The production of secondary metabolites of the food fermentation fungi Monascus spp. was found to be influenced by different chemical and physical signals, such as nutrients, osmolarity, pH and light (Miyake et al., 2005; Lee et al., 2006; Babitha et al., 2007). It is widely accepted that heterotrimeric G-protein signalling pathways play a pivotal role in perceiving and transmitting KU-60019 order many of the external signals to elicit specific responses in cells, including regulating the production of metabolites (Calvo et al., 2002; Yu, 2006). The deletion of Mga1 in M. ruber M7 resulted in an selleck increase in the production of citrinin and pigments,

providing genetic evidence that the signalling pathway mediated by the Gα-subunit encoded by Mga1 is involved in the regulation of production of secondary metabolites in Monascus spp. Monascus metabolites, for example red pigments and monacolins, are widely used as natural food colorants or antihypercholesterolemic agents, but citrinin is nephrotoxic in mammalian systems. To prevent the negative effects of citrinin, scientific work has been carried out to identify low- or non-citrinin-producing Monascus strains (Chen & Hu, 2005; Wang et al., 2005; Chen et al., 2008a; Pattanagul from et al., 2008). Some results have shown that citrinin was detectable in strains of M. ruber (Wang et al., 2005; Pattanagul et al., 2008), whereas other results revealed that M. ruber was not a citrinin producer, as functional citrinin biosynthesis genes, such as polyketide synthase gene pksCT, were absent in M. ruber (Chen et al., 2008a). However, the strain used in our study, M. ruber M7, produced citrinin both in YES (this study)

and in steamed rice media (Chen & Hu, 2005). The most extensively studied G-protein signalling model in filamentous fungi is A. nidulans. Intensive analysis of the A. nidulans genome has been carried out, and more than 40 genes/putative genes were predicted to encode components that function in G-protein signalling pathways (Lafon et al., 2006; Yu, 2006). A proposed model of the roles of these signalling proteins in controlling A. nidulans growth, development and secondary metabolism has been described (Yu, 2006). As signal perception and signal processing via the G-protein signalling pathway are complex processes, identification of one component of this pathway is not enough to shed light on a possible regulation mechanism.

, 1985), different virulence factors (Mekalanos, BIBF 1120 1992; Bajaj et al., 1996) and several other genes (Weber et al., 2006; Gunasekera et al., 2008). Hitherto, most of the well-characterized osmoregulated genes correspond to genes that are upregulated following an osmotic upshift (Cairney et al., 1985; Han et al., 2005; Weber et al., 2006; Gunasekera et al., 2008). Nevertheless, adaptation

to low-osmolarity conditions must also result in regulation of genes that are specifically required to cope with these conditions. In this work we designed a genetic strategy focused on identifying genes that are optimally expressed at low osmolarity in Salmonella enterica serovar Typhimurium (S. Typhimurium). We report here the identification of a novel LysR-type transcriptional regulator (LTTR) that shows osmolarity-dependent expression. Bacterial strains, plasmids and phages used are listed in Table 1. Cells were routinely grown in Luria–Bertani (LB) medium. For some experiments, LB was modified by adding NaCl up to 0.5 M (LB 0.5 M NaCl) or by not including NaCl (LB 0 M NaCl). When required, X-Gal (40 μg mL−1) was added to the culture medium. Antibiotics were used at the following concentrations: kanamycin

(Km) 50 μg mL−1; ampicillin (Ap) 25 and 50 μg mL−1; tetracycline (Tc) 15 μg mL−1. The growth temperature was 37 °C unless noted otherwise. To obtain phage-free isolates, transductants were purified by streaking on EBU plates (LB agar PD0325901 molecular weight supplemented with 0.25% glucose, 0.25% KH2PO4, 12.5 mg L−1 Evans Blue and 25 mg L−1

fluorescein). Restriction digestion, ligation, transformation, agarose gel electrophoresis and DNA manipulations were performed using standard procedures. For plasmid DNA preparations, the Wizard® Plus SV Minipreps kit (Promega) was used. DNA was recovered Hydroxychloroquine clinical trial from agarose gels by electroelution or Qiaquick® gel extraction kit (Qiagen). The Wizard® Clean-Up System (Promega) was used for purification of DNA fragments. PCR experiments were performed in the Perkin Elmer GeneAmp PCR System 2400 according to standard protocols, using DynaZyme™ (Finnzyme). Oligonucleotides used are listed in Table 1. DNA sequencing reactions were carried out according to the instructions of the BigDye® Terminator v3.1 Cycle Sequencing Kit from Applied Biosystems. A lysate of P22HTint4 phage grown on S. Typhimurium strain TT10288 (hisD9953∷MudJ) was used to transduce strain TT1704, selecting Km resistance (Kmr). The recipient strain carried the nontransducible deletion his-9953, which avoids homologous recombination with MudJ from donor lysate. To identify the gene in which MudJ was inserted, Sau3A-partially digested TT1704-OS chromosomal DNA was ligated with BglII-digested cosmid pLA2917. The ligation was packed following instructions from Gigapack III (Stratagene) and used to infect E. coli HB101.

, 1985), different virulence factors (Mekalanos, signaling pathway 1992; Bajaj et al., 1996) and several other genes (Weber et al., 2006; Gunasekera et al., 2008). Hitherto, most of the well-characterized osmoregulated genes correspond to genes that are upregulated following an osmotic upshift (Cairney et al., 1985; Han et al., 2005; Weber et al., 2006; Gunasekera et al., 2008). Nevertheless, adaptation

to low-osmolarity conditions must also result in regulation of genes that are specifically required to cope with these conditions. In this work we designed a genetic strategy focused on identifying genes that are optimally expressed at low osmolarity in Salmonella enterica serovar Typhimurium (S. Typhimurium). We report here the identification of a novel LysR-type transcriptional regulator (LTTR) that shows osmolarity-dependent expression. Bacterial strains, plasmids and phages used are listed in Table 1. Cells were routinely grown in Luria–Bertani (LB) medium. For some experiments, LB was modified by adding NaCl up to 0.5 M (LB 0.5 M NaCl) or by not including NaCl (LB 0 M NaCl). When required, X-Gal (40 μg mL−1) was added to the culture medium. Antibiotics were used at the following concentrations: kanamycin

(Km) 50 μg mL−1; ampicillin (Ap) 25 and 50 μg mL−1; tetracycline (Tc) 15 μg mL−1. The growth temperature was 37 °C unless noted otherwise. To obtain phage-free isolates, transductants were purified by streaking on EBU plates (LB agar SB203580 cost supplemented with 0.25% glucose, 0.25% KH2PO4, 12.5 mg L−1 Evans Blue and 25 mg L−1

fluorescein). Restriction digestion, ligation, transformation, agarose gel electrophoresis and DNA manipulations were performed using standard procedures. For plasmid DNA preparations, the Wizard® Plus SV Minipreps kit (Promega) was used. DNA was recovered Amoxicillin from agarose gels by electroelution or Qiaquick® gel extraction kit (Qiagen). The Wizard® Clean-Up System (Promega) was used for purification of DNA fragments. PCR experiments were performed in the Perkin Elmer GeneAmp PCR System 2400 according to standard protocols, using DynaZyme™ (Finnzyme). Oligonucleotides used are listed in Table 1. DNA sequencing reactions were carried out according to the instructions of the BigDye® Terminator v3.1 Cycle Sequencing Kit from Applied Biosystems. A lysate of P22HTint4 phage grown on S. Typhimurium strain TT10288 (hisD9953∷MudJ) was used to transduce strain TT1704, selecting Km resistance (Kmr). The recipient strain carried the nontransducible deletion his-9953, which avoids homologous recombination with MudJ from donor lysate. To identify the gene in which MudJ was inserted, Sau3A-partially digested TT1704-OS chromosomal DNA was ligated with BglII-digested cosmid pLA2917. The ligation was packed following instructions from Gigapack III (Stratagene) and used to infect E. coli HB101.

g. the obligate methanotroph Methylocystis parvus OBBP and the facultative methanotroph Methylocystis strain H2s). Comparison of the genomes of obligate and facultative methanotrophs with those of facultative methylotrophs could also prove useful in this endeavor. In addition, proteomic and/or metabolomic strategies could be applied to help deduce metabolic pathway(s) used for uptake of multicarbon compounds. Other important questions that remain to be answered include: 1

What environmental conditions promote obligate vs. facultative methanotrophy? Finally, it is interesting to note that not only Methylocystis strains are found in many different environments, but also Methylocella strains. Members of both genera are widely distributed throughout the globe, found not only in peat bogs, but also in acidic forest and arctic soils as well as environments with pH values >7.0 (Bowman, Apoptosis inhibitor 2006; Dedysh, 2009; Rahman et al., 2011). Such findings indicate that facultative methanotrophy may be widespread. “
“The opportunistic human fungal pathogen Candida glabrata is closely related to Saccharomyces cerevisiae, yet it has evolved to survive within mammalian hosts. Which traits help C. glabrata to adapt to this different environment? Which specific

responses are crucial for its survival selleckchem in the host? The main differences seem to include an extended repertoire of adhesin genes, high drug resistance, an enhanced ability to sustain prolonged starvation and adaptations of the transcriptional wiring of key stress response genes. Here, we discuss the properties of C. glabrata with a focus on the differences to related fungi. “
“A growing interest in culturable diversity has required

microbiologists to think seriously about microbial preservation. In addition to the isolation and cultivation of pure strains, adequate preservation without changes in morphological, physiological and however genetic traits is necessary. This review consolidates different methods used for preservation of microorganisms with an emphasis on cryopreservation and lyophilization. The critical points of cryopreservation and lyophilization are highlighted to explain how several extrinsic and intrinsic factors affect the cell survival and recovery during the process of long-term preservation. Factors responsible for alteration in genotypic and phenotypic integrity of cultures during preservation and methods used for their evaluation have been incorporated. We emphasize the importance of depositories and highlight their current funding status. Future areas for preservation research, including cell dormancy, ecosystem and community level preservation and the effects of the viable but non-culturable state on post-preservation recovery of the cells are also discussed.

However, altogether these results indicate that YFP-MinDEc likely recognizes the same lipid spirals as GFP-MinDBs (Barák et al., 2008). Although no apparent phenotypical effect of MinEEc expression on B. subtilis cells was observed, its localization was also inspected. The fluorescence signal was dispersed through the cytoplasm. Only a few spots near cell poles were visible, which can be caused by inclusion body formation (not shown). However, the immunoblot analysis revealed only minimal degradation of the fusion protein (Fig. 3c). These data indicate that MinEEc-GFP

is probably unable to co-operate with B. subtilis Min proteins. To further study MinDEc functioning in B. subtilis, find more we also examined three previously undescribed mutant forms possessing mutations in different parts of the molecule (G209D, S89P and I23N). The cell lengths were measured in B. subtilis strains IB1135, IB1136 and IB1137, which express GFP-MinDEc(G209D), GFP-MinDEc(S89P)

and GFP-MinDEc(I23N), respectively, from the amyE locus under the control of Pxyl. The ability of mutant versions of MinDEc to substitute MinDBs in ΔminD cells and their localization pattern was tested as above for the GFP-MinDEc. Interestingly, one of these mutants, GFP-MinDEc(G209D), showed different effects on B. subtilis selleck chemical cells in comparison with GFP-MinDEc. This protein was not able to elongate wild-type B. subtilis cells. Moreover, it did not suppress the minicell phenotype of ΔminDBs cells at a lower concentration as was shown for the nonmutated version of GFP-MinDEc (Table 2). However, the GFP-MinDEc(G209D) fluorescence pattern was not perturbed and resembled YFP-MinDEc localization (Fig. this website 4c). Despite the homology between Min proteins in Gram-negative and Gram-positive bacteria, two different paths of their action have been observed and thus two models have been proposed. In E. coli the Min system behaves extremely dynamically. An oscillatory movement of the Min proteins on helical trajectories was described (Shih et al., 2003, 2005). By contrast, in B. subtilis a static localization

of Min proteins at the division sites and at the cell poles was observed (Edwards & Errington, 1997; Marston et al., 1998). We have recently shown that GFP-MinDBs, attracted to negatively charged phospholipids, localizes to the membrane in helical structures (Barák et al., 2008). In this study the functioning and localization of E. coli Min proteins in B. subtilis cells was determined. MinCEc and also YFP-MinDEc cause elongation of B. subtilis cells, indicating that they are functional and are able to cause division delay or to block the cell division. However, MinCEc was not able to repair defects caused by minCBs disruption. In this case we are not able to exclude the possibility of a negative effect of minCBs deletion on expression of minDBs.

Only two patients in the combined NVP arm and two patients in the ATZ/r arm of the study experienced cardiac disorders of division of acquired immunodeficiency syndrome (DAIDS) grade 3 or 4. In the combined NVP arm, one patient experienced angina pectoris and one patient http://www.selleckchem.com/products/azd4547.html experienced myopericarditis. In the ATZ/r arm, one patient experienced MI, and another experienced cardiac failure. Primary data from the ARTEN study confirm that the favourable virological and immunological responses to NVP combined with TDF/FTC are maintained through 48 weeks of treatment and are noninferior

to those of ATZ/r [in combination with the same dual nucleoside reverse transcriptase inhibitor (NRTI) backbone] with a similar safety profile [23]. The data presented here also suggest a more favourable lipid profile with NVP than with ATZ/r when combined with TDF/FTC. There are many risk factors for CVD. Known factors include smoking, being overweight, lack of exercise, insulin resistance, elevated waist circumference, hypertension, elevated LDL-c, elevated triglycerides and low HDL-c. For HIV-infected patients receiving treatment with ARVs, the risk of CVD may be significantly greater than in the general population [27]. Increased levels of TG, TC and LDL-c, reduced levels of HDL-c, unfavourable changes in the TC:HDL-c ratio and lipodystrophy are common side effects in patients receiving certain ARV drugs

[1–4]. The cardiac disorders of DAIDS grade 3 or 4 reported in four patients in the ARTEN study (two in each arm) probably relate to pre-existing cardiovascular Oxymatrine risk factors, although a role of antiretroviral therapy (ART) cannot be ruled Carfilzomib out. With respect to serum lipid levels, traditionally LDL-c is recognized as the primary target of cholesterol-lowering therapy. However, full evaluation of lipid-related risk (i.e. TC, HDL-c, the TC:HDL-c ratio and TG levels) should

also be considered, as these measures play an important role as markers of cardiovascular risk [28]. Although ATZ/r use was associated with markedly lower LDL-c increases compared with NVP, LDL-c is known to be an incomplete measure of atherogenic lipoproteins because very low-density lipoprotein (VLDL) remnants are also likely to contribute to coronary heart disease [29]. In contrast, ApoB measurement includes all atherogenic lipoproteins, with each VLDL and LDL particle having one molecule of ApoB, making ApoB a more reliable measure of the concentration of proatherogenic particles [30]. The Apolipoprotein-related Mortality Risk (AMORIS) study showed that elevated ApoB levels were strongly related to increased cardiovascular risk and were also a stronger marker of cardiovascular risk than LDL-c [31]. In the current study, NVP-containing regimens showed no difference in ApoB, significantly greater increases in HDL-c and ApoA1, and an improved ApoB:ApoA1 ratio over 48 weeks compared with the ATZ/r regimen.

22 μm) glucose–nitrate (100 mg L−1 NO3-N) solution to yield a final find protocol C : N ratio of 40 : 1 so that the ectomycorrhizal fungi were not C limited (Fransson et al., 2007). Discs (3 mm diameter) of fungal inoculum were cut from actively growing fungal mycelia and once mycelium had projected around the plugs, they were transferred to the serum bottles (three discs per bottle, one fungus per bottle; n=10 for each fungus). A control treatment (without fungal inoculum; n=10) was also established. All treatments were incubated in the dark as static, aerobic cultures at 20 °C. The total

growth period was 14 days. A short growth period was used here, which is atypical of ectomycorrhizal fungal incubation experiments, because fungal N2O production is often not prolonged (Bleakley & Tiedje, 1982). After the first 3 days, the headspace in each bottle was sealed and reduced to 10% v/v O2 by replacing with sterile helium www.selleckchem.com/products/GDC-0980-RG7422.html gas and an injection of 10 mL sterile O2 into the headspace. A concentration of 10% v/v was selected based on data from a preliminary experiment under initially aerobic conditions, which showed no detectable N2O production over 32 days where headspace O2 concentrations declined to ∼14% v/v (Prendergast-Miller,

2009; unpublished data). After an additional 24 h under low O2 conditions (day 1), the headspace gas concentrations were analysed: N2O and carbon dioxide (CO2) were determined on an Agilent 6890 gas chromatograph, fitted with an ECD FID and methanizer, and O2 was measured using a MAP Test 800 O2-meter. Fungal mycelium was collected and dried for 48 h at 60 °C for fungal biomass Forskolin chemical structure determination. The nitrate concentration and pH of the growth medium were also analysed (n=5 for each treatment). The remaining bottles (n=5 for each treatment) were sampled similarly after a further 10 days of growth (day 10). Differences between and within treatments in gas production, fungal biomass and media nitrate and pH analyses were compared using one-way anovas and paired t-tests with minitab (v. 15). The ectomycorrhizal fungi formed

a mycelial mat over the liquid surface, whereas F. lichenicola formed a globular submerged culture. Fungal biomass was measured twice, 24 h after 10% v/v O2 conditions had been induced (day 1) and after a further 10 days of growth (day 10) (Fig. 1). Growth occurred in all three species from the initial biomass to day 1 (P<0.05). During the low O2 period, no significant increase in biomass occurred in T. fibrillosa or F. lichenicola, although P. involutus biomass showed a small, but not significant increase (P=0.053). The ectomycorrhizal fungi P. involutus and T. fibrillosa produced more total biomass over the experimental period (P<0.05) than F. lichenicola, reflecting the preferential growth medium for ectomycorrhizal fungi. After 24 h under low O2 conditions (day 1; ∼10% v/v O2, no significant difference between treatments), no N2O was detected from any treatment (limit of detection ∼0.

As shown in Fig. 1a, the colony size of strain Δpeps was considerably smaller than that of strain JM101 on M9 agar medium. When cell growth SCH772984 nmr was monitored in flask cultivations, strain Δpeps did not grow in M9 medium (Fig. 1b). This growth deficiency was substantially restored by supplementing casamino acids in M9 medium. We then examined the effect of dipeptide addition on cell growth of strain Δpeps on M9 agar plate. As a result, it was revealed that several dipeptides, including Ala-Gln, inhibited the growth of strain Δpeps. Among these dipeptides,

we chose Ala-Gln and glycyl-l-tyrosine (Gly-Tyr), the structure of which is rather different from Ala-Gln. When Ala-Gln or Gly-Tyr was added to M9 agar medium, colony formation was not observed in strain Δpeps (Fig. 1a). In contrast, strain JM101 could grow

under the same condition by degrading dipeptides to amino acids. These results indicate that Ala-Gln or Gly-Tyr themselves, not their component amino acids, have an inhibitory effect on a multiple peptidase-deficient BMS 907351 E. coli. Because Ala-Gln addition was inhibitory on strain Δpeps, it was expected that active efflux of Ala-Gln was mediated by a family of transmembrane proteins referred to as multidrug-efflux transporters. Therefore, we transformed strain Δpeps with plasmids carrying one of the 34 coding sequences assumed to be a multidrug-efflux transporter gene in E. coli and examined the effect of their overexpression on the growth of strain Δpeps in the presence of

either Ala-Gln or Gly-Tyr (Fig. 2a). Of these 34 genes, bcr, norE, ydeE and yeeO conferred resistance to Ala-Gln or Gly-Tyr. In contrast, overexpression of acrAB, emrAB, emrE, emrKY, marRAB, rhtA, rhtB, rhtC, yajR, ybjG, yceE, yceL, ydeA, ydeD, ydhC, yeaS, yedA, yegB, yfiK, yfiS, ygaZ, ygeD, yggA, yidY, yieO, yjeH, yjiO, ykuC, ymtF or yrgJ genes had no influence on the very growth of strain Δpeps (data not shown). Accordingly, the four genes were chosen as candidates for dipeptide transporters. Table 2 lists the four multidrug-efflux transporter genes being considered as dipeptide transporter candidates. To further examine the effect of overexpression of four multidrug-efflux transporter genes selected by dipeptide resistance, cell growth was monitored in flask cultivation. Growth of strain Δpeps was defective in M9 glucose liquid medium even with no addition of dipeptides (Fig. 1b). There are two possibilities considered as the cause of the hampered growth of strain Δpeps. One is the reduced availability of intracellular amino acids derived from protein degradation due to the loss of peptidase activity. This is partially true because the addition of casamino acids to M9 medium significantly improved the growth of strain Δpeps (Fig. 1b). However, this effect seemed to be general because the same result was obtained in the parental strain JM101.

Six and four of the 10 experts participating in the study were from European and non-European countries, respectively. Eight of the experts declared the use of one to four rule-based expert systems while two declared the use of none. Figure 1 shows the predictions made by the 10 experts and by the EuResist engine for each of the individual TCEs. Overall, 15 of the 25 TCEs met the criteria for definition of virological success. The EuResist engine mislabelled six cases; three successes and three see more failures (accuracy 0.76). The mean±SD number of incorrect calls made by the human experts was 9.1±1.9 (mean±SD accuracy 0.64±0.07), with only one expert making

the same number of errors as EuResist and all the others making more (range 8–13). Overall, Smad inhibitor there were apparently more failures mislabelled as successes than the opposite (mean±SD 5.3±2.7 vs. 3.8±1.6, respectively) but the difference was not significant and reflected the uneven distribution of failures and successes in the data set (Table 2). Also, European and non-European experts did not differ in their performance (mean±SD number of wrong calls 9.8±1.7 vs. 8.0±1.6, respectively), nor did they

show different use of the expert systems. There was no correlation between the number of expert systems consulted and the number of errors made. When ROC analysis was applied to determine the sensitivity and specificity of prediction of treatment success, EuResist was found to be not significantly better than the mean prediction computed by the human experts, nor was it better than any of the individual experts (Fig. 2). The only significant difference in performance was between the best and worst experts, as measured by the area under the ROC curve (P=0.011). The agreement among the experts in terms of binary classification of success and failure was only fair, as revealed by the relatively low kappa multirater agreement Atazanavir value (0.355). There were only five (20%) cases where all the experts made the same prediction. In all of these, the outcome was as predicted and the EuResist system prediction agreed with

the opinion of the experts. The mean±SD coefficient of variation for the quantitative prediction made by the experts for the individual TCEs was also relatively high (55.9±22.4%). However, the significant correlation between the quantitative prediction generated by EuResist and the average quantitative prediction provided by the experts showed a strong positive relationship (Pearson r=0.695, P<0.0001), with considerable inter-individual variation. According to the Bland–Altman plot (Fig. 3), the difference between the quantitative predictions given by the experts and by the EuResist engine is independent of the mean of the two values, indicating that there was no systematic error related to the magnitude of the predicted probability. A closer look at the individual TCEs revealed four cases where the EuResist engine as well as eight or nine of the human experts made incorrect calls.