Differentiation of glioblastoma from solitary brain metastasis using deep ensembles: Empirical estimation of uncertainty for clinical reliability.

BACKGROUND AND OBJECTIVES: To develop a clinically reliable deep learning model to differentiate glioblastoma (GBM) from solitary brain metastasis (SBM) by providing predictive uncertainty estimates and interpretability. METHODS: A total of 469 patients (300 GBM, 169 SBM) were enrolled in the institutional training set. Deep ensembles based on DenseNet121 were trained on multiparametric MRI. The model performance was validated in the external test set consisting of 143 patients (101 GBM, 42 SBM). Entropy values for each input were evaluated for uncertainty measurement; based on entropy values, the datasets were split to high- and low-uncertainty groups. In addition, entropy values of out-of-distribution (OOD) data from unknown class (257 patients with meningioma) were compared to assess uncertainty estimates of the model. The model interpretability was further evaluated by localization accuracy of the model. RESULTS: On external test set, the area under the curve (AUC), accuracy, sensitivity and specificity of the deep ensembles were 0.83 (95 % confidence interval [CI] 0.76-0.90), 76.2 %, 54.8 % and 85.2 %, respectively. The performance was higher in the low-uncertainty group than in the high-uncertainty group, with AUCs of 0.91 (95 % CI 0.83-0.98) and 0.58 (95 % CI 0.44-0.71), indicating that assessment of uncertainty with entropy values ascertained reliable prediction in the low-uncertainty group. Further, deep ensembles classified a high proportion (90.7 %) of predictions on OOD data to be uncertain, showing robustness in dataset shift. Interpretability evaluated by localization accuracy provided further reliability in the "low-uncertainty and high-localization accuracy" subgroup, with an AUC of 0.98 (95 % CI 0.95-1.00). CONCLUSIONS: Empirical assessment of uncertainty and interpretability in deep ensembles provides evidence for the robustness of prediction, offering a clinically reliable model in differentiating GBM from SBM.

Design of high-performance entangling logic in silicon quantum dot systems with Bayesian optimization.

Device engineering based on computer-aided simulations is essential to make silicon (Si) quantum bits (qubits) be competitive to commercial platforms based on superconductors and trapped ions. Combining device simulations with the Bayesian optimization (BO), here we propose a systematic design approach that is quite useful to procure fast and precise entangling operations of qubits encoded to electron spins in electrode-driven Si quantum dot (QD) systems. For a target problem of the controlled-X (CNOT) logic operation, we employ BO with the Gaussian process regression to evolve design factors of a Si double QD system to the ones that are optimal in terms of speed and fidelity of a CNOT logic driven by a single microwave pulse. The design framework not only clearly contributes to cost-efficient securing of solutions that enhance performance of the target quantum operation, but can be extended to implement more complicated logics with Si QD structures in experimentally unprecedented ways.

Large-scale duplication events underpin population-level flexibility in tRNA gene copy number in Pseudomonas fluorescens SBW25.

The complement of tRNA genes within a genome is typically considered to be a (relatively) stable characteristic of an organism. Here, we demonstrate that bacterial tRNA gene set composition can be more flexible than previously appreciated, particularly regarding tRNA gene copy number. We report the high-rate occurrence of spontaneous, large-scale, tandem duplication events in laboratory populations of the bacterium Pseudomonas fluorescens SBW25. The identified duplications are up to ∼1 Mb in size (∼15% of the wildtype genome) and are predicted to change the copy number of up to 917 genes, including several tRNA genes. The observed duplications are inherently unstable: they occur, and are subsequently lost, at extremely high rates. We propose that this unusually plastic type of mutation provides a mechanism by which tRNA gene set diversity can be rapidly generated, while simultaneously preserving the underlying tRNA gene set in the absence of continued selection. That is, if a tRNA set variant provides no fitness advantage, then high-rate segregation of the duplication ensures the maintenance of the original tRNA gene set. However, if a tRNA gene set variant is beneficial, the underlying duplication fragment(s) may persist for longer and provide raw material for further, more stable, evolutionary change.

Machine-learning-based diagnosis of thyroid fine-needle aspiration biopsy synergistically by Papanicolaou staining and refractive index distribution.

We developed a machine learning algorithm (MLA) that can classify human thyroid cell clusters by exploiting both Papanicolaou staining and intrinsic refractive index (RI) as correlative imaging contrasts and evaluated the effects of this combination on diagnostic performance. Thyroid fine-needle aspiration biopsy (FNAB) specimens were analyzed using correlative optical diffraction tomography, which can simultaneously measure both, the color brightfield of Papanicolaou staining and three-dimensional RI distribution. The MLA was designed to classify benign and malignant cell clusters using color images, RI images, or both. We included 1535 thyroid cell clusters (benign: malignancy = 1128:407) from 124 patients. Accuracies of MLA classifiers using color images, RI images, and both were 98.0%, 98.0%, and 100%, respectively. As information for classification, the nucleus size was mainly used in the color image; however, detailed morphological information of the nucleus was also used in the RI image. We demonstrate that the present MLA and correlative FNAB imaging approach has the potential for diagnosing thyroid cancer, and complementary information from color and RI images can improve the performance of the MLA.

ROCK1 regulates insulin secretion from ß-cells.

OBJECTIVE: The endocrine pancreatic ß-cells play a pivotal role in maintaining whole-body glucose homeostasis and its dysregulation is a consistent feature in all forms of diabetes. However, knowledge of intracellular regulators that modulate ß-cell function remains incomplete. We investigated the physiological role of ROCK1 in the regulation of insulin secretion and glucose homeostasis. METHODS: Mice lacking ROCK1 in pancreatic ß-cells (RIP-Cre; ROCK1loxP/loxP, ß-ROCK1-/-) were studied. Glucose and insulin tolerance tests as well as glucose-stimulated insulin secretion (GSIS) were measured. An insulin secretion response to a direct glucose or pyruvate or pyruvate kinase (PK) activator stimulation in isolated islets from ß-ROCK1-/- mice or ß-cell lines with knockdown of ROCK1 was also evaluated. A proximity ligation assay was performed to determine the physical interactions between PK and ROCK1. RESULTS: Mice with a deficiency of ROCK1 in pancreatic ß-cells exhibited significantly increased blood glucose levels and reduced serum insulin without changes in body weight. Interestingly, ß-ROCK1-/- mice displayed a progressive impairment of glucose tolerance while maintaining insulin sensitivity mostly due to impaired GSIS. Consistently, GSIS markedly decreased in ROCK1-deficient islets and ROCK1 knockdown INS-1 cells. Concurrently, ROCK1 blockade led to a significant decrease in intracellular calcium and ATP levels and oxygen consumption rates in isolated islets and INS-1 cells. Treatment of ROCK1-deficient islets or ROCK1 knockdown ß-cells either with pyruvate or a PK activator rescued the impaired GSIS. Mechanistically, we observed that glucose stimulation in ß-cells greatly enhanced ROCK1 binding to PK. CONCLUSIONS: Our findings demonstrate that ß-cell ROCK1 is essential for glucose-stimulated insulin secretion and for glucose homeostasis and that ROCK1 acts as an upstream regulator of glycolytic pyruvate kinase signaling.

Outer Membrane Porin F in E. coli Is Critical for Effective Predation by Bdellovibrio.

Bdellovibrio and like organisms (BALOs) are a unique bacterial group that live by predating on other bacteria, consuming them from within to grow and replicate before the progeny come out to complete the life cycle. The mechanisms by which these predators recognize their prey and differentiate them from nonprey bacteria, however, are still not clear. Through genetic knockout and complementation studies in different Escherichia coli strains, we found that Bdellovibrio bacteriovorus strain 109J recognizes outer membrane porin F (OmpF) on the E. coli surface and that the activity of the E. coli EnvZ-OmpR regulatory system significantly impacts predation kinetics. OmpF is not the only signal by which BALOs recognize their prey, however, as B. bacteriovorus could eventually predate on the E. coli ΔompF mutant after prolonged incubation. Furthermore, recognizing OmpF as a prey surface structure was dependent on the prey strain, as knocking out OmpF protein homologues in other prey species, including Escherichia fergusonii, Klebsiella pneumoniae, and Salmonella enterica, did not always reduce the predation rate. Consequently, although OmpF was found to be an important surface component used by Bdellovibrio to efficiently recognize and attack E. coli, future work is needed to determine what other prey surface structures are recognized by these predators. IMPORTANCE Bdellovibrio bacteriovorus and like organisms (BALOs) are Gram-negative predatory bacteria that attack other Gram-negative bacteria by penetrating their periplasm and consuming them from within to obtain the nutrients necessary for the predator's growth and replication. How these predators recognize their prey, however, has remained a mystery. Here, we show that the outer membrane porin F (OmpF) in E. coli is recognized by B. bacteriovorus strain 109J and that the loss of this protein leads to severely delayed predation. However, predation of several other prey species was not dependent on the recognition of this protein or its homologues, indicating that there are other structures recognized by the predators on the prey surface that are yet to be discovered.

Minimax Optimal Bandits for Heavy Tail Rewards.

Stochastic multiarmed bandits (stochastic MABs) are a problem of sequential decision-making with noisy rewards, where an agent sequentially chooses actions under unknown reward distributions to minimize cumulative regret. The majority of prior works on stochastic MABs assume that the reward distribution of each action has bounded supports or follows light-tailed distribution, i.e., sub-Gaussian distribution. However, in a variety of decision-making problems, the reward distributions follow a heavy-tailed distribution. In this regard, we consider stochastic MABs with heavy-tailed rewards, whose p th moment is bounded by a constant νp for . First, we provide theoretical analysis on sub-optimality of the existing exploration methods for heavy-tailed rewards where it has been proven that existing exploration methods do not guarantee a minimax optimal regret bound. Second, to achieve the minimax optimality under heavy-tailed rewards, we propose a minimax optimal robust upper confidence bound (MR-UCB) by providing tight confidence bound of a p -robust estimator. Furthermore, we also propose a minimax optimal robust adaptively perturbed exploration (MR-APE) which is a randomized version of MR-UCB. In particular, unlike the existing robust exploration methods, both proposed methods have no dependence on νp . Third, we provide the gap-dependent and independent regret bounds of proposed methods and prove that both methods guarantee the minimax optimal regret bound for a heavy-tailed stochastic MAB problem. The proposed methods are the first algorithm that theoretically guarantees the minimax optimality under heavy-tailed reward settings to the best of our knowledge. Finally, we demonstrate the superiority of the proposed methods in simulation with Pareto and Fréchet noises with respect to regrets.

The Kiss of Death: Serratia marcescens Antibacterial Activities against Staphylococcus aureus Requires Both de novo Prodigiosin Synthesis and Direct Contact.

Prodigiosin possesses antibacterial activities, but as a highly hydrophobic compound, it raised the question about how Serratia marcescens introduce this compound to other microbes. Here, we demonstrate that the production of prodigiosin by newly isolated S. marcescens RH10 correlates with its antibacterial activity against a multidrug-resistant strain of S. aureus, with this pathogen's viability decreasing 6-log over 24 h. While S. marcescens RH10 does secrete membrane vesicles that carry prodigiosin, this antibiotic was not active in this form, with 5 mg/L prodigiosin leading to only a 1.22-fold reduction in the S. aureus viability while the same quantity of purified prodigiosin led to a 2800-fold reduction. Contact assays, however, showed increased activity, with a 3-log loss in the S. aureus viabilities in only 6 h as long as de novo production of prodigiosin occurred. The role of prodigiosin was confirmed further by generating an isogenic ΔpigA mutant in S. marcescens RH10, based on the draft genome sequence reported here, to inhibit the synthesis of prodigiosin. In all experiments performed, this mutant was unable to kill S. aureus. Finally, the possibility that the type VI secretion system present in S. marcescens may also be important was also explored as it is known to be used by this strain to kill other microbes. The results here, however, found no obvious activity against S. aureus. In conclusion, the results presented here show prodigiosin requires both cell-to-cell contact and de novo synthesis for it to be effective as an antibiotic for its native host. IMPORTANCE The antibacterial activities of prodigiosin are well-established but, as a hydrophobic molecule, the mechanisms used to introduce it to susceptible microbes has never been studied. We found here, in contrast to violacein, another hydrophobic antibiotic that can be transferred using membrane vesicles (MVs), prodigiosin is also carried from Serratia marcescens in MVs released but its resulting activities were severely mitigated compared to the freely added compound, suggesting it is more tightly bound to the MVs than violacein. This led us to hypothesize that cell-to-cell contact is needed, which we demonstrate here. As well, we show de novo synthesis of prodigiosin is needed for it to be effective. As violacein- and prodigiosin-producing bacterial strains are both beneficial to amphibians, where they help protect the skin against pathogens, the findings presented here provide an important ecological perspective as they show the mechanisms used differ according to the antibacterial produced.

Biotechnological Activities and Applications of Bacterial Pigments Violacein and Prodigiosin.

In this review, we discuss violacein and prodigiosin, two chromogenic bacterial secondary metabolites that have diverse biological activities. Although both compounds were "discovered" more than seven decades ago, interest into their biological applications has grown in the last two decades, particularly driven by their antimicrobial and anticancer properties. These topics will be discussed in the first half of this review. The latter half delves into the current efforts of groups to produce these two compounds. This includes in both their native bacterial hosts and heterogeneously in other bacterial hosts, including discussing some of the caveats related to the yields reported in the literature, and some of the synthetic biology techniques employed in this pursuit.

Evaluation of conditional treatment effects of adjuvant treatments on patients with synovial sarcoma using Bayesian subgroup analysis.

BACKGROUND: The impact of adjuvant chemotherapy or radiation therapy on the survival of patients with synovial sarcoma (SS), which is a rare soft-tissue sarcoma, remains controversial. Bayesian statistical approaches and propensity score matching can be employed to infer treatment effects using observational data. Thus, this study aimed to identify the individual treatment effects of adjuvant therapies on the overall survival of SS patients and recognize subgroups of patients who can benefit from specific treatments using Bayesian subgroup analyses. METHODS: We analyzed data from patients with SS obtained from the surveillance, epidemiology, and end results (SEER) public database. These data were collected between 1984 and 2014. The treatment effects of chemotherapy and radiation therapy on overall survival were evaluated using propensity score matching. Subgroups that could benefit from radiation therapy or chemotherapy were identified using Bayesian subgroup analyses. RESULTS: Based on a stratified Kaplan-Meier curve, chemotherapy exhibited a positive average causal effect on survival in patients with SS, whereas radiation therapy did not. The optimal subgroup for chemotherapy includes the following covariates: older than 20 years, male, large tumor (longest diameter > 5 cm), advanced stage (SEER 3), extremity location, and spindle cell type. The optimal subgroup for radiation therapy includes the following covariates: older than 20 years, male, large tumor (longest diameter > 5 cm), early stage (SEER 1), extremity location, and biphasic type. CONCLUSION: In this study, we identified high-risk patients whose variables include age (age > 20 years), gender, tumor size, tumor location, and poor prognosis without adjuvant treatment. Radiation therapy should be considered in the early stages for high-risk patients with biphasic types. Conversely, chemotherapy should be considered for late-stage high-risk SS patients with spindle cell types.

Chromobacterium violaceum delivers violacein, a hydrophobic antibiotic, to other microbes in membrane vesicles.

This study describes Chromobacterium violaceum's use of extracellular membrane vesicles (MVs) to both solubilize and transport violacein to other microorganisms. Violacein is a hydrophobic bisindole with known antibiotic activities against other microorganisms. Characterization of the MVs found they carried more violacein than protein (1.37 ± 0.19-fold), suggesting they may act as a reservoir for this compound. However, MVs are not produced in response to violacein - a ΔvioA isogenic mutant, which is incapable of making violacein, actually produced significantly more MVs (3.2-fold) than the wild-type strain. Although violacein is insoluble in water (Log Poctanol:water = 3.34), 79.5% remained in the aqueous phase when it was present within the C. violaceum MVs, an increase in solubility of 1740-fold. Moreover, tests with a strain of Staphylococcus aureus showed MV-associated violacein is bactericidal, with 3.1 mg/l killing 90% of S. aureus in 6 h. Tests with the ΔvioA MVs found no loss in the S. aureus viability, even when its MVs were added at much higher concentrations, demonstrating violacein is the active component within the wild-type MVs. In conclusion, our study clearly demonstrates C. violaceum produces MVs and uses them as vehicles to solubilize violacein and transport this hydrophobic antibiotic to other microbes.

Control of type O foot-and-mouth disease by vaccination in Korea, 2014-2015.

On December 3, 2014, a type O foot-and-mouth disease (FMD) outbreak began in Korea. Although vaccinations were administered, FMD cases increased steadily for five months, and reached 185 cases by April 2015. Most of the affected animals were pigs, which are vulnerable to vaccination. The FMD virus belonged to the South-East Asia (SEA) topotype that had been observed three times in Korea between April 2010 and July 2014. However, the FMD virus isolated in December 2014 had a unique feature; that is, partial deletion of the 5´ non-coding region, a deletion not seen in previous SEA topotype isolates identified in Korea. We conclude that this outbreak included the introduction of a new FMD strain to Korea, and that Korea was now affected by genetically similar FMD virus strains that are related to those from neighboring countries.

Characterization of Echinostoma cinetorchis endoribonuclease, RNase H.

Echinostoma cinetorchis is an oriental intestinal fluke causing significant pathological damage to the small intestine. The aim of this study was to determine a full-length cDNA sequence of E. cinetorchis endoribonuclease (RNase H; EcRNH) and to elucidate its molecular biological characters. EcRNH consisted of 308 amino acids and showed low similarity to endoribonucleases of other parasites (<40%). EcRNH had an active site centered on a putative DDEED motif instead of DEDD conserved in other species. A recombinant EcRNH produced as a soluble form in Escherichia coli showed enzymatic activity to cleave the 3'-O-P bond of RNA in a DNA-RNA duplex, producing 3'-hydroxyl and 5'-phosphate. These findings may contribute to develop antisense oligonucleotides which could damage echinostomes and other flukes.

Development of a polymerase chain reaction applicable to rapid and sensitive detection of Clonorchis sinensis eggs in human stool samples.

Microscopic examination of eggs of parasitic helminths in stool samples has been the most widely used classical diagnostic method for infections, but tiny and low numbers of eggs in stool samples often hamper diagnosis of helminthic infections with classical microscopic examination. Moreover, it is also difficult to differentiate parasite eggs by the classical method, if they have similar morphological characteristics. In this study, we developed a rapid and sensitive polymerase chain reaction (PCR)-based molecular diagnostic method for detection of Clonorchis sinensis eggs in stool samples. Nine primers were designed based on the long-terminal repeat (LTR) of C. sinensis retrotransposon1 (CsRn1) gene, and seven PCR primer sets were paired. Polymerase chain reaction with each primer pair produced specific amplicons for C. sinensis, but not for other trematodes including Metagonimus yokogawai and Paragonimus westermani. Particularly, three primer sets were able to detect 10 C. sinensis eggs and were applicable to amplify specific amplicons from DNA samples purified from stool of C. sinensis-infected patients. This PCR method could be useful for diagnosis of C. sinensis infections in human stool samples with a high level of specificity and sensitivity.

Identification and characterization of a mitochondrial manganese superoxide dismutase of Spirometra erinacei.

A gene encoding the manganese superoxide dismutase (Mn-SOD) of Spirometra erinacei was identified, and the biochemical properties of the recombinant enzyme were partially characterized. The S. erinacei Mn-SOD gene consisted of 669 bp, which encoded 222 amino acids. A sequence analysis of the gene showed that it had typical molecular structures, including characteristic metal-binding residues and motifs that were conserved in Mn-SODs. An analysis of the N-terminal presequence of S. erinacei Mn-SOD revealed that it had physiochemical characteristics commonly found in mitochondria-targeting sequences and predicted that the enzyme is located in the mitochondria. A biochemical analysis also revealed that the enzyme is a typical Mn-SOD. The enzyme was consistently expressed in both S. erinacei plerocercoid larvae and adult worms. Our results collectively suggested that S. erinacei Mn-SOD is a typical mitochondrial Mn-SOD and may play an important role in parasite physiology, detoxifying excess superoxide radicals generated in the mitochondria.