Unlike normal cells, cancer cells frequently exhibit supernumerary centrosomes, leading to formation of multipolar spindles that can trigger cell death. Nevertheless, cancer cells with supernumerary centrosomes escape the deadly consequences of unequal segregation of genomic material by coalescing their centrosomes into two poles. This unique trait of cancer cells presents a promising target for cancer therapy, focusing on selectively attacking cells with supernumerary centrosomes. Nek2A is a kinase involved in mitotic regulation, including the centrosome cycle, where it phosphorylates linker proteins to separate centrosomes. In this study, we investigated if Nek2A also prevents clustering of supernumerary centrosomes, akin to its separation function. Reduction of Nek2A activity, achieved through knockout, silencing, or inhibition, promotes centrosome clustering, whereas its overexpression results in inhibition of clustering. Significantly, prevention of centrosome clustering induces cell death, but only in cancer cells with supernumerary centrosomes, both in vitro and in vivo. Notably, none of the known centrosomal (e.g., CNAP1, Rootletin, Gas2L1) or non-centrosomal (e.g., TRF1, HEC1) Nek2A targets were implicated in this machinery. Additionally, Nek2A operated via a pathway distinct from other proteins involved in centrosome clustering mechanisms, like HSET and NuMA. Through TurboID proximity labeling analysis, we identified novel proteins associated with the centrosome or microtubules, expanding the known interaction partners of Nek2A. KIF2C, in particular, emerged as a novel interactor, confirmed through coimmunoprecipitation and localization analysis. The silencing of KIF2C diminished the impact of Nek2A on centrosome clustering and rescued cell viability. Additionally, elevated Nek2A levels were indicative of better patient outcomes, specifically in those predicted to have excess centrosomes. Therefore, while Nek2A is a proposed target, its use must be specifically adapted to the broader cellular context, especially considering centrosome amplification. Discovering partners such as KIF2C offers fresh insights into cancer biology and new possibilities for targeted treatment.

Background: Glioblastoma is the most common and aggressive primary brain tumor with extremely poor prognosis, highlighting an urgent need for developing novel treatment options. Identifying epigenetic vulnerabilities of cancer cells can provide excellent therapeutic intervention points for various types of cancers.

Method: In this study, we investigated epigenetic regulators of glioblastoma cell survival through CRISPR/Cas9 based genetic ablation screens using a customized sgRNA library EpiDoKOL, which targets critical functional domains of chromatin modifiers.

Results: Screens conducted in multiple cell lines revealed ASH2L, a histone lysine methyltransferase complex subunit, as a major regulator of glioblastoma cell viability. ASH2L depletion led to cell cycle arrest and apoptosis. RNA sequenc‐ ing and greenCUT&RUN together identified a set of cell cycle regulatory genes, such as TRA2B, BARD1, KIF20B, ARID4A and SMARCC1 that were downregulated upon ASH2L depletion. Mass spectrometry analysis revealed the interaction partners of ASH2L in glioblastoma cell lines as SET1/MLL family members including SETD1A, SETD1B, MLL1 and MLL2. We further showed that glioblastoma cells had a differential dependency on expression of SET1/MLL family members for survival. The growth of ASH2L‐depleted glioblastoma cells was markedly slower than controls in orthotopic in vivo models. TCGA analysis showed high ASH2L expression in glioblastoma compared to low grade gliomas and immuno‐ histochemical analysis revealed significant ASH2L expression in glioblastoma tissues, attesting to its clinical relevance. Therefore, high throughput, robust and affordable screens with focused libraries, such as EpiDoKOL, holds great prom‐ ise to enable rapid discovery of novel epigenetic regulators of cancer cell survival, such as ASH2L.

Conclusion: Together, we suggest that targeting ASH2L could serve as a new therapeutic opportunity for glioblastoma.

Background: In many applications of bioinformatics, data stem from distinct het- erogeneous sources. One of the well-known examples is the identification of drug–target interactions (DTIs), which is of significant importance in drug discovery. In this paper, we propose a novel framework, manifold optimization based kernel preserving embedding (MOKPE), to efficiently solve the problem of modeling heterogeneous data. Our model projects heterogeneous drug and target data into a unified embedding space by preserving drug–target interactions and drug–drug, target–target similarities simultaneously.

Results: We performed ten replications of ten-fold cross validation on four different drug–target interaction network data sets for predicting DTIs for previously unseen drugs. The classification evaluation metrics showed better or comparable performance compared to previous similarity-based state-of-the-art methods. We also evaluated MOKPE on predicting unknown DTIs of a given network. Our implementation of the proposed algorithm in R together with the scripts that replicate the reported experiments is publicly available at https://github.com/ocbinatli/mokpe.

Background: Cefiderocol is generally active against carbapenem-resistant Klebsiella spp. (CRK) with higher MICs against metallo-beta-lactamase producers. There is a variation in cefiderocol interpretive criteria determined by EUCAST and CLSI. Our objective was to test CRK isolates against cefiderocol and compare cefiderocol susceptibilities using EUCAST and CLSI interpretive criteria.

Methods: A unique collection (n = 254) of mainly OXA-48-like- or NDM-producing CRK bloodstream isolates were tested against cefiderocol with disc diffusion (Mast Diagnostics, UK). Beta-lactam resistance genes and multilocus sequence types were identified using bioinformatics analyses on complete bacterial genomes.

Results: Median cefiderocol inhibition zone diameter was 24 mm (interquartile range [IQR] 24–26 mm) for all isolates and 18 mm (IQR 15–21 mm) for NDM producers. We observed significant variability between cefiderocol susceptibilities using EUCAST and CLSI breakpoints, such that 26% and 2% of all isolates, and 81% and 12% of the NDM producers were resistant to cefiderocol using EUCAST and CLSI interpretive criteria, respectively.

Conclusions: Cefiderocol resistance rates among NDM producers are high using EUCAST criteria. Breakpoint variability may have significant implications on patient outcomes. Until more clinical outcome data are available, we suggest using EUCAST interpretive criteria for cefiderocol susceptibility testing.

Purpose: Tocilizumab, a monoclonal IL-6 receptor blocker, is an effective agent for severe-to-critical cases of COVID-19; however, its target patients for the optimum use need to be detailed. We performed a systematic review and meta-analysis to define its effect among severely ill but non-intubated cases with COVID-19.

Methods: We searched PubMed, Scopus, Web of Science, MEDLINE, Cochrane Central Register of Controlled Trials (CENTRAL), Medrxiv, and Biorxiv until February 13, 2022, for non-intubated cases, and included randomized-controlled trials (RCT) based on bias assessment. The primary outcomes were the requirement of invasive mechanical ventilation and mortality. Random effect and fixed-effect models were used. The heterogeneity was measured using the χ² and I² statistics, with χ² p ≤ 0.05 and I² ≥ 50% indicating the presence of significant heterogeneity. We registered the study to the International Prospective Register of Systematic Reviews (PROSPERO) with the registration number CRD42021232575.

Results: Among 261 articles, 11 RCTs were included. The pooled analysis of the 11 RCTs demonstrated that the rate of mortality was significantly lower in the tocilizumab group than in the control group (20.0% and 24.2%, OR: 0.84, 95% CI 0.73–0.96, and heterogeneity I² = 0%. p = 0.82.). The mechanical ventilation rate was lower in the tocilizumab group than the control group (27% vs 35.2%, OR: 0.76, 95% CI 0.67–0.86, and heterogeneity I² = 6%. p = 0.39).

Conclusion: Among non-intubated severe COVID-19 cases, tocilizumab reduces the risk of invasive mechanical ventilation and mortality compared to standard-of-care treatment.

Introduction: Aminoglycosides are used for the treatment of carbapenemase-producing Klebsiella pneumoniae (CPK) infections. 16S rRNA methyltransferases (RMTs) confer resistance to all aminoglycosides and are often cocarried with NDM.

Hypothesis/Gap Statement: There is a dart of studies looking at the aminoglycoside resistance mechanisms for invasive CPK isolates, particularly in OXA-48 endemic settings.

Aim: We aimed to determine the prevalence of RMTs and their association with beta lactamases and MLSTs amongst aminoglycoside-resistant CPK bloodstream isolates in an OXA-48 endemic setting.

Methodology: CPK isolates (n=181), collected as part of a multicentre cohort study, were tested for amikacin, gentamicin and tobramycin susceptibility using custom-made sensititre plates (GN2XF, Thermo Fisher Scientific). All isolates were previously subjected to whole-genome sequencing. Carbapenemases, RMTs, MLSTs and plasmid incompatibility groups were detected on the assembled genomes.

Results: Of the 181 isolates, 109 (60 %) were resistant to all three aminoglycosides, and 96 of 109 (88 %) aminoglycoside-resistant isolates carried an RMT (85 ArmA, 10 RmtC, 4 RmtF1; three isolates cocarried ArmA and RmtC). Main clonal types associated with ArmA were ST2096 (49/85, 58 %) and ST14 (24/85, 28 %), harbouring mainly OXA-232 and OXA-48 +NDM, respectively. RmtC was cocarried with NDM (5/10) on ST395, and NDM +OXA-48 or NDM +KPC (4/10) on ST14, ST15 and ST16. All RMT producers also carried CTX-M-15, and the majority cocarried SHV-106, TEM-150 and multiple other antibiotic resistance genes. The majority of the isolates harboured a combination of IncFIB, IncH and IncL/M type plasmids. Non-NDM producing isolates remained susceptible to ceftazidime-avibactam.

Conclusion: Aminoglycoside resistance amongst CPK bloodstream isolates is extremely common and mainly driven by clonal spread of ArmA carried on ST2096 and ST14, associated with OXA-232 and OXA48 +NDM carriage, respectively.

MicroRNA (miRNA) alterations significantly impact the formation and progression of human cancers. miRNAs interact with messenger RNAs (mRNAs) to facilitate degradation or translational repression. Thus, identifying miRNA–mRNA regulatory modules in cohorts of primary tumor tissues are fundamental for understanding the biology of tumor heterogeneity and precise diagnosis and treatment. We established a multitask learning sparse regularized factor regression (MSRFR) method to determine key tissue- and cohort-specific miRNA–mRNA regulatory modules from expression profiles of tumors. MSRFR simultaneously models the sparse relationship between miRNAs and mRNAs and extracts tissue- and cohort-specific miRNA–mRNA regulatory modules separately. We tested the model’s ability to determine cohort-specific regulatory modules of multiple cancer cohorts from the same tissue and their underlying tissue-specific regulatory modules by extracting similarities between cancer cohorts (i.e., blood, kidney, and lung). We also detected tissue-specific and cohort-specific signatures in the corresponding regulatory modules by comparing our findings from various other tissues. We show that MSRFR effectively determines cancer-related miRNAs in cohort-specific regulatory modules, distinguishes tissue- and cohort-specific regulatory modules from each other, and extracts tissue-specific information from different cohorts of disease-related tissue. Our findings indicate that the MSRFR model can support current efforts in precision medicine to define tumor-specific miRNA–mRNA signatures.

The impact of COVID-19 across the United States (US) has been heterogeneous, with rapid spread and greater mortality in some areas compared with others. We used geographically-linked data to test the hypothesis that the risk for COVID-19 was defined by location and sought to define which demographic features were most closely associated with elevated COVID-19 spread and mortality. We leveraged geographically-restricted social, economic, political, and demographic information from US counties to develop a computational framework using structured Gaussian process to predict county-level case and death counts during the pandemic’s initial and nationwide phases. After identifying the most predictive information sources by location, we applied an unsupervised clustering algorithm and topic modeling to identify groups of features most closely associated with COVID-19 spread. Our model successfully predicted COVID-19 case counts of unseen locations after examining case counts and demographic information of neighboring locations, with overall Pearson’s correlation coefficient and the proportion of variance explained as 0.96 and 0.84 during the initial phase and 0.95 and 0.87 during the nationwide phase, respectively. Aside from population metrics, presidential vote margin was the most consistently selected spatial feature in our COVID-19 prediction models. Urbanicity and 2020 presidential vote margins were more predictive than other demographic features. Models trained using death counts showed similar performance metrics. Topic modeling showed that counties with similar socioeconomic and demographic features tended to group together, and some of these feature sets were associated with COVID-19 dynamics. Clustering of counties based on these feature groups found by topic modeling revealed groups of counties that experienced markedly different COVID-19 spread. We conclude that topic modeling can be used to group similar features and identify counties with similar features in epidemiologic research.

Multitask multiple kernel learning (MKL) algorithms combine the capabilities of incorporating different data sources into the prediction model and using the data from one task to improve the accuracy on others. However, these methods do not necessarily produce interpretable results. Restricting the solutions to the set of interpretable solutions increases the computational burden of the learning problem significantly, leading to computationally prohibitive run times for some important biomedical applications. That is why we propose a multitask MKL formulation with a clustering of tasks and develop a highly time-efficient solution approach for it. Our solution method is based on the Benders decomposition and treating the clustering problem as finding a given number of tree structures in a graph; hence, it is called the forest formulation. We use our method to discriminate early-stage and late-stage cancers using genomic data and gene sets and compare our algorithm against two other algorithms. The two other algorithms are based on different approaches for linearization of the problem while all algorithms make use of the cutting-plane method. Our results indicate that as the number of tasks and/or the number of desired clusters increase, the forest formulation becomes increasingly favorable in terms of computational performance.

Ceftazidime-avibactam exhibits good in vitro activity against carbapenem resistant Klebsiella carrying OXA-48-like enzymes. We tested two hundred unique carbapenem resistant Klebsiella blood stream isolates (71% with single OXA-48-like carbapenemases, including OXA-48, n = 62; OXA-232, n = 57; OXA-244, n = 17; OXA-181, n = 5) that were collected as part of a multicentre study against ceftazidime-avibactam using Etest (bioMérieux, Marcyl’Étoile, France), 10/4 μg disc (Thermo Fisher) and Sensititre Gram Negative EURGNCOL Plates (Lyophilized panels, Sensititre, Thermo Fisher) with the aim of comparing the performances of the Etest and disc to that of Sensititre. Ceftazidime-avibactam MIC50/90 was 2/>16 mg/L for the entire collection and was 2/4 mg/L for single OXA-48-like producers. Categorical and essential agreements between the Etest and Sensititre were 100% and 97%, respectively. Categorical agreement between the disc and Sensititre was 100%. Etest and 10/4 μg discs are suitable alternatives to Sensititre for ceftazidime-avibactam sensitivity testing for OXA-48-like producers.

Motivation: Dataset sizes in computational biology have been increased drastically with the help of improved data collection tools and increasing size of patient cohorts. Previous kernel-based machine learning algorithms proposed for increased interpretability started to fail with large sample sizes, owing to their lack of scalability. To overcome this problem, we proposed a fast and efficient multiple kernel learning (MKL) algorithm to be particularly used with large-scale data that integrates kernel approximation and group Lasso formulations into a conjoint model. Our method extracts significant and meaningful information from the genomic data while conjointly learning a model for out-of-sample prediction. It is scalable with increasing sample size by approximating instead of calculating distinct kernel matrices.

Results: To test our computational framework, namely, Multiple Approximate Kernel Learning (MAKL), we demonstrated our experiments on three cancer datasets and showed that MAKL is capable to outperform the baseline algorithm while using only a small fraction of the input features. We also reported selection frequencies of approximated kernel matrices associated with feature subsets (i.e. gene sets/pathways), which helps to see their relevance for the given classification task. Our fast and interpretable MKL algorithm producing sparse solutions is promising for computational biology applications considering its scalability and highly correlated structure of genomic datasets, and it can be used to discover new biomarkers and new therapeutic guidelines.

Objective: To develop machine learning (ML) models that predict postoperative remission, remission at last visit, and resis- tance to somatostatin receptor ligands (SRL) in patients with acromegaly and to determine the clinical features associated with the prognosis.

Methods: We studied outcomes using the area under the receiver operating characteristics (AUROC) values, which were reported as the performance metric. To determine the importance of each feature and easy interpretation, Shapley Additive explanations (SHAP) values, which help explain the outputs of ML models, are used.

Results: One-hundred fifty-two patients with acromegaly were included in the final analysis. The mean AUROC values resulting from 100 independent replications were 0.728 for postoperative 3 months remission status classification, 0.879 for remission at last visit classification, and 0.753 for SRL resistance status classification. Extreme gradient boosting model demonstrated that preoperative growth hormone (GH) level, age at operation, and preoperative tumor size were the most important predictors for early remission; resistance to SRL and preoperative tumor size represented the most important pre- dictors of remission at last visit, and postoperative 3-month insulin-like growth factor 1 (IGF1) and GH levels (random and nadir) together with the sparsely granulated somatotroph adenoma subtype served as the most important predictors of SRL resistance.

Conclusions: ML models may serve as valuable tools in the prediction of remission and SRL resistance.

A prospective, multicentre observational cohort study of carbapenem-resistant Klebsiella spp. (CRK) bloodstream infections was conducted in Turkey from June 2018 to June 2019. One hundred eighty-seven patients were recruited. Single OXA-48-like carbapenemases predominated (75%), followed by OXA-48-like/NDM coproducers (16%). OXA-232 constituted 31% of all OXA-48-like carbapenemases and was mainly carried on ST2096. Thirty-day mortality was 44% overall and 51% for ST2096. In the multivariate cox regression analysis, SOFA score and immunosuppression were significant predictors of 30-day mortality and ST2096 had a non-significant effect. All OXA-48-like producers remained susceptible to ceftazidime-avibactam.

We aimed to describe the effect of aminoglycosides and tigecycline to reduce the mortality in colistin- and carbapenem-resistant Klebsiella pneumoniae (ColR-CR-Kp) infections. We included the studies with defined outcomes after active or non-active antibiotic treatment of ColR-CR-Kp infections. The active treatment was defined as adequate antibiotic use for at least 3 days (72 h) after the diagnosis of ColR-CR-Kp infection by culture. The Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement and the checklist of PRISMA 2020 was applied. Crude and adjusted odds ratios (OR) with 95% confidence interval (CI) were calculated and pooled in the random effects model. Adding aminoglycosides to the existing treatment regimen reduced overall mortality significantly (OR 0.34, 95% CI 0.20–0.58). Overall mortality was 34% in patients treated with aminoglycoside-combined regimens and was 60% in patients treated with non-aminoglycoside regimens. Treatment with tigecycline is not found to reduce mortality (OR: 0.76, 95% CI: 0.47–1.23). Our results suggest that aminoglycoside addition to the existing regimen of colistin- and carbapenem-resistant Klebsiella pneumoniae infections reduces mortality significantly.

This paper describes the elimination of healthcare associated Acinetobacter baumannii infections in a highly endemic region. A prospective, observational study was performed between October 2012 and October 2017. Acinetobacter baumannii was isolated from 59 patients, and >95% similarity was demonstrated among isolates of seven patients (DiversiLab™, BioMérieux). Carbapenemase activity was detected in 15 out of 17 (88%) isolates, and all were OXA-23 type. The control of Acinetobacter baumannii outbreaks can be achieved by the close follow-up supported by molecular techniques, strict application of infection control measures, and isolation of the transferred patients.

We performed a systematic review and meta-analysis for the effectiveness of Favipiravir on the fatality and the requirement of mechanical ventilation for the treatment of moderate to severe COVID-19 patients. We searched available literature and reported it by using PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. Until June 1, 2021, we searched PubMed, bioRxiv, medRxiv, ClinicalTrials.gov, Cochrane Central Register of Controlled Trials (CENTRAL), and Google Scholar by using the keywords “Favipiravir” and terms synonymous with COVID-19. Studies for Favipiravir treatment compared to standard of care among moderate and severe COVID-19 patients were included. Risk of bias assessment was performed using Revised Cochrane risk of bias tool for randomized trials (RoB 2) and ROBINS-I assessment tool for non-randomized studies. We defined the outcome measures as fatality and requirement for mechanical ventilation. A total of 2702 studies were identified and 12 clinical trials with 1636 patients were analyzed. Nine out of 12 studies were randomized controlled trials. Among the randomized studies, one study has low risk of bias, six studies have moderate risk of bias, and 2 studies have high risk of bias. Observational studies were identified as having moderate risk of bias and non-randomized study was found to have serious risk of bias. Our meta-analysis did not reveal any significant difference between the intervention and the comparator on fatality rate (OR 1.11, 95% CI 0.64–1.94) and mechanical ventilation requirement (OR 0.50, 95% CI 0.13–1.95). There is no significant difference in fatality rate and mechanical ventilation requirement between Favipiravir treatment and the standard of care in moderate and severe COVID-19 patients.

Background: Identification of molecular mechanisms that determine tumour progression in cancer patients is a prerequisite for developing new disease treatment guidelines. Even though the predictive performance of current machine learning models is promising, extracting significant and meaningful knowledge from the data simultaneously during the learning process is a difficult task considering the high-dimensional and highly correlated nature of genomic datasets. Thus, there is a need for models that not only predict tumour volume from gene expression data of patients but also use prior information coming from pathway/gene sets during the learning process, to distinguish molecular mechanisms which play crucial role in tumour progression and therefore, disease prognosis.

Results: In this study, instead of initially choosing several pathways/gene sets from an available set and training a model on this previously chosen subset of genomic features, we built a novel machine learning algorithm, PrognosiT, that accomplishes both tasks together. We tested our algorithm on thyroid carcinoma patients using gene expression profiles and cancer-specific pathways/gene sets. Predictive performance of our novel multiple kernel learning algorithm (PrognosiT) was comparable or even better than random forest (RF) and support vector regression (SVR). It is also notable that, to predict tumour volume, PrognosiT used gene expression features less than one-tenth of what RF and SVR algorithms used.

Conclusions: PrognosiT was able to obtain comparable or even better predictive performance than SVR and RF. Moreover, we demonstrated that during the learning process, our algorithm managed to extract relevant and meaningful pathway/gene sets information related to the studied cancer type, which provides insights about its progression and aggressiveness. We also compared gene expressions of the selected genes by our algorithm in tumour and normal tissues, and we then discussed up- and down-regulated genes selected by our algorithm while learning, which could be beneficial for determining new biomarkers.

Background: The prevalence of SARS-CoV-2 infection among health care workers (HCWs) provide information to for the spread of COVID-19 within health care facilities, and the risk groups.

Objectives: We aimed to describe the rate of SARS-CoV-2 seroprevalence and its determinants among HCWs.

Data sources: We used Web of Science, PubMed, Scopus, MEDLINE, EBSCOhost and Cochrane Library.

Study eligibility criteria: We included the reports of SARS-CoV-2 seroprevalence with a sample size of minimum 1000 HCWs.

Methods: The study was registered at the International Prospective Register of Systematic Reviews (PROSPERO, no: CRD42021230456). We used PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement. The keywords were “COVID-19”, “SARS-CoV-2”, “Coronavirus”, “seroprevalence”, “health care workers” and “risk factors”.

Results: In total 4329 reports were retrieved, the duplications were removed; after filtering according to the title and abstract, 25 studies were selected. Risk of bias was assessed in 25 studies; it was low in 13 studies, medium in four studies, and high in eight studies. In meta-analysis by using the random effect model, the weighted average of seroprevalence was calculated as 8% (95% CI: 6-10%). The pooled seroprevalence rates of the selected variables that have a rate over the average were male HCWs with 9% (95% CI: 7-11%); HCWs from ethnic minorities with 13% (95% CI: 9-17%); high exposure 9% (95% CI: 6-13%); exposure to the virus outside the health care setting 22% (95% CI: 14-32%).

Conclusions: Our analysis indicate a SARS-CoV-2 seroprevalence rate of 8% among studies included >1000 HCWs for the year 2020, before vaccinations started. The most common risk factors associated with higher seroprevalence rate were ethnicity, male gender and having higher number of household contacts. Working as a frontline HCW was inconsistent in its association with higher seroprevalence.

Objective: To compare the effectiveness of different mask types in limiting the dispersal of coughed air.

Method: The Schlieren method with a single curved mirror was used in this study. Coughed air has a slightly higher temperature than ambient air, which generates a refractive index gradient. A curved mirror with a radius of curvature of 10 m and a diameter of 60 cm was used. The spread of the cough wavefront was investigated among five subjects wearing: (1) no mask; (2) a single surgical mask; (3) a double surgical mask; (4) a cloth mask; (5) a valveless N95 mask; and (6) a valved N95 mask.

Results: All mask types reduced the size of the contaminated region significantly. The percentage reduction in the cross-sectional area of the contaminated region for the same mask types on different subjects revealed by normalized data suggests that the fit of a mask plays an important role.

Conclusions: No significant difference in the spread of coughed air was found between the use of a single surgical mask or a double surgical mask. Cloth masks may be effective, depending on the quality of the cloth. Valved N95 masks exclusively protect the user. The fit of a mask is an important factor to minimize the contaminated region.

We described the significance of systematic monitoring nationwide antimicrobial stewardship programs (ASPs) in primary care. All the prescriptions given by family physicians were recorded in Prescription Information System established by the Turkish Medicines and Medical Devices Agency of Ministry of Health. We calculated, for each prescription, “antibiotics amount” as number of boxes times number of items per box for medicines that belong to antiinfectives for systemic use (i.e., J01 block in the Anatomical Therapeutic Chemical Classification System). We compared the antibiotics amount before (2015) and after (2016) the extensive training programs for the family physicians. We included 266,389,209 prescriptions from state-operated family healthcare units (FHUs) between January 1, 2015 and December 31, 2016. These prescriptions were given by 26,313 individual family physicians in 22,518 FHUs for 50,713,181 individual patients. At least one antimicrobial was given in 37,024,232 (28.31%) prescriptions in 2015 and 36,154,684 (26.66%) prescriptions in 2016. The most common diagnosis was “acute upper respiratory infections (AURI)” (i.e., J00-J06 block in the 10th revision of the International Statistical Classification of Diseases and Related Health Problems) with 28.05%. The average antibiotics amount over prescriptions with AURI decreased in 79 out of 81 provinces, and overall rate of decrease in average antibiotics amount was 8.33%, where 28 and 53 provinces experienced decreases (range is between 28.63% and −3.05%) above and below this value, respectively. In the most successful province, the highest decrease in average amount of “other beta-lactam antibacterials” per prescription for AURI was 49.63% in January. Computational analyses on a big data set collected from a nationwide healthcare system brought a significant contribution in improving ASPs.

We proposed the hypothesis that high-risk clones of colistin-resistant K. pneumoniae (ColR-Kp) possesses a high number of virulence factors and has enhanced survival capacity against the neutrophil activity. We studied virulence genes of ColR-Kp isolates and neutrophil response in 142 patients with invasive ColR-Kp infections. The ST101 and ST395 ColR-Kp infections had higher 30-day mortality (58%, p = 0.005 and 75%, p = 0.003). The presence of yersiniabactin biosynthesis gene (ybtS) and ferric uptake operon associated gene (kfu) were significantly higher in ST101 (99%, p ≤ 0.001) and ST395 (94%, p < 0.012). Being in ICU (OR: 7.9; CI: 1.43–55.98; p = 0.024), kfu (OR:27.0; CI: 5.67–179.65; p < 0.001) and ST101 (OR: 17.2; CI: 2.45–350.40; p = 0.01) were found to be predictors of 30-day mortality. Even the neutrophil uptake of kfu+-ybtS+ ColR-Kp was significantly higher than kfu–ybtS- ColR-Kp (phagocytosis rate: 78% vs. 65%, p < 0.001), and the kfu+-ybtS+ ColR-Kp survived more than kfu–ybtS- ColR-Kp (median survival index: 7.90 vs. 4.22; p = 0.001). The kfu+-ybtS+ ColR-Kp stimulated excessive NET formation. Iron uptake systems in high-risk clones of colistin-resistant K. pneumoniae enhance the success of survival against the neutrophil phagocytic defense and stimulate excessive NET formation. The drugs targeted to iron uptake systems would be a promising approach for the treatment of colistin-resistant high-risk clones of K. pneumoniae infections.

Recently, it has been showed that cancer missense mutations selectively target the neighborhood of hinge residues, which are key sites in protein dynamics. Here, we show that this approach can be extended to find previously unknown candidate mutations and genes. To this aim, we developed a computational pipeline to detect significantly enriched three-dimensional (3D) clustering of missense mutations around hinge residues. The hinge residues were detected by applying a Gaussian network model. By systematically analyzing the PanCancer compendium of somatic mutations in nearly 10000 tumors from the Cancer Genome Atlas, we identified candidate genes and mutations in addition to well known ones. For instance, we found significantly enriched 3D clustering of missense mutations in known cancer genes including CDK4, CDKN2A, TCL1A, and MAPK1. Beside these known genes, we also identified significantly enriched 3D clustering of missense mutations around hinge residues in PLA2G4A, which may lead to excessive phosphorylation of the extracellular signal-regulated kinases. Furthermore, we demonstrated that hinge-based features improves pathogenicity prediction for missense mutations. Our results show that the consideration of clustering around hinge residues can help us explain the functional role of the mutations in known cancer genes and identify candidate genes.

Background: There is limited evidence on the modification or stopping of antiretroviral therapy (ART) regimens, including novel antiretroviral drugs. The aim of this study was to evaluate the discontinuation of first ART before and after the availability of better tolerated and less complex regimens by comparing the frequency, reasons and associations with patient characteristics.

Methods: A total of 3019 ART-naive patients registered in the HIV-TR cohort who started ART between Jan 2011 and Feb 2017 were studied. Only the first modification within the first year of treatment for each patient was included in the analyses. Reasons were classified as listed in the coded form in the web-based database. Cumulative incidences were analysed using competing risk function and factors associated with discontinuation of the ART regimen were examined using Cox proportional hazards models and Fine-Gray competing risk regression models.

Results: The initial ART regimen was discontinued in 351 out of 3019 eligible patients (11.6%) within the first year. The main reason for discontinuation was intolerance/toxicity (45.0%), followed by treatment simplification (9.7%), patient willingness (7.4%), poor compliance (7.1%), prevention of future toxicities (6.0%), virologic failure (5.4%), and provider preference (5.4%). Non-nucleoside reverse transcriptase inhibitor (NNRTI)-based (aHR = 4.4, [95% CI 3.0–6.4]; p < 0.0001) or protease inhibitor (PI)-based regimens (aHR = 4.3, [95% CI 3.1–6.0]; p < 0.0001) relative to integrase strand transfer inhibitor (InSTI)-based regimens were significantly associated with ART discontinuation. ART initiated at a later period (2015-Feb 2017) (aHR = 0.6, [95% CI 0.4–0.9]; p < 0.0001) was less likely to be discontinued. A lower rate of treatment discontinuation for intolerance/toxicity was observed with InSTI-based regimens (2.0%) than with NNRTI- (6.6%) and PI-based regimens (7.5%) (p < 0.001). The percentage of patients who achieved HIV RNA < 200 copies/mL within 12 months of ART initiation was 91% in the ART discontinued group vs. 94% in the continued group (p > 0.05).

Conclusion: ART discontinuation due to intolerance/toxicity and virologic failure decreased over time. InSTI-based regimens were less likely to be discontinued than PI- and NNRTI-based ART.

Keywords: Antiretroviral therapy, Treatment modification, Integrase strand transfer inhibitor, Treatment outcome, Cohort study

Objectives: The case fatality rate (CFR) of coronavirus disease 2019 (COVID-19) varies significantly between countries. We aimed to describe the associations of health indicators with the national CFRs of COVID-19.

Methods: We identified health for each country indicators potentially associated with the national CFRs of COVID-19. We extracted data for 18 variables from international administrative data sources for 34 member countries of the Organization for Economic Cooperation and Development (OECD). We excluded the collinear variables and examined the 16 variables in multivariable analysis. A dynamic web-based model was developed to analyse and display the associations for the CFRs of COVID-19. We followed the Guideline for Accurate and Transparent Health Estimates Reporting (GATHER).

Results: In multivariable analysis, the variables significantly associated with the increased CFRs were percent of obesity in ages >18 years (β = 3.26, 95% CI = [1.20, 5.33], p = 0.003), tuberculosis incidence (β = 3.15, 95% CI = [1.09, 5.22], p = 0.004), duration (days) since first death due to COVID-19 (β = 2.89, 95% CI = [0.83, 4.96], p = 0.008), median age (β = 2.83, 95% CI = [0.76, 4.89], p = 0.009). The COVID-19 test rate (β = -3.54, 95% CI = [-5.60, -1.47], p = 0.002), hospital bed density (β = -2.47, 95% CI = [-4.54, -0.41], p = 0.021), and rural population ratio (β = -2.19, 95% CI = [-4.25, -0.13], p = 0.039) decreased the CFR.

Conclusions: The pandemic hits population-dense cities. Available hospital beds should be increased. Test capacity should be increased to enable more effective diagnostic tests. Older patients and patients with obesity and their caregivers should be warned about a potentially increased risk.

Motivation: Micro RNAs (miRNAs) are known as the important components of RNA silencing and post-transcriptional gene regulation, and they interact with messenger RNAs (mRNAs) either by degradation or by translational repression. miRNA alterations have a significant impact on the formation and progression of human cancers. Accordingly, it is important to establish computational methods with high predictive performance to identify cancer-specific miRNA–mRNA regulatory modules.

Results: We presented a two-step framework to model miRNA–mRNA relationships and identify cancer-specific modules between miRNA and mRNA from their matched expression profiles of more than 9000 primary tumors. We first estimated the regulatory matrix between miRNA and mRNA expression profiles by solving multiple linear programming problems. We then formulated a unified regularized factor regression (RFR) model that simultaneously estimates the effective number of modules (i.e. latent factors) and extracts modules by decomposing regulatory matrix into two low-rank matrices. Our RFR model groups correlated miRNAs together and correlated mRNAs together, and also controls sparsity levels of both matrices. These attributes lead to interpretable results with high predictive performance. We applied our method on a very comprehensive data collection including 32 TCGA cancer types. To find the biological relevance of our approach, we performed functional gene set enrichment and survival analyses. A large portion of the identified modules are significantly enriched in Hallmark, PID and KEGG pathways/gene sets. To validate the identified modules, we also performed literature validation as well as validation using experimentally supported miRTarBase database.

Motivation: Genomic information is increasingly being used in diagnosis, prognosis and treatment of cancer. The severity of the disease is usually measured by the tumor stage. Therefore, identifying pathways playing an important role in progression of the disease stage is of great interest. Given that there are similarities in the underlying mechanisms of different cancers, in addition to the considerable correlation in the genomic data, there is a need for machine learning methods that can take these aspects of genomic data into account. Furthermore, using machine learning for studying multiple cancer cohorts together with a collection of molecular pathways creates an opportunity for knowledge extraction.

Results: We studied the problem of discriminating early- and late-stage tumors of several cancers using genomic information while enforcing interpretability on the solutions. To this end, we developed a multitask multiple kernel learning (MTMKL) method with a co-clustering step based on a cutting-plane algorithm to identify the relationships between the input tasks and kernels. We tested our algorithm on 15 cancer cohorts and observed that, in most cases, MTMKL outperforms other algorithms (including random forests, support vector machine and single-task multiple kernel learning) in terms of predictive power. Using the aggregate results from multiple replications, we also derived similarity matrices between cancer cohorts, which are, in many cases, in agreement with available relationships reported in the relevant literature.

Androgen receptor (AR) signalling is essential in nearly all prostate cancers. Any alterations to AR-mediated transcription can have a profound effect on carcinogenesis and tumour growth. While mutations of the AR protein have been extensively studied, little is known about those somatic mutations that occur at the non-coding regions where AR binds DNA. Using clinical whole genome sequencing, we show that AR binding sites have a dramatically increased rate of mutations that is greater than any other transcription factor and specific to only prostate cancer. Demonstrating this may be common to lineage-specific transcription factors, estrogen receptor binding sites were also found to have elevated rate of mutations in breast cancer. We provide evidence that these mutations at AR binding sites, and likely other related transcription factors, are caused by faulty repair of abasic sites. Overall, this work demonstrates that non-coding AR binding sites are frequently mutated in prostate cancer and can impact enhancer activity.

Objectives: We aimed to develop a prospective prediction tool on Crimean–Congo haemorrhagic fever (CCHF) to identify geographic regions at risk. The tool could support public health decision makers in implementation of an effective control strategy in a timely manner.

Methods: We used monthly surveillance data between 2004 and 2015 to predict case counts between 2016 and 2017 prospectively. Turkish nationwide surveillance dataset collected by Ministry of Health contained 10,411 confirmed CCHF cases. We collected potential explanatory covariates about climate, land use, and animal and human population at risk to capture spatiotemporal transmission dynamics. We developed a structured Gaussian process algorithm and prospectively tested this tool predicting the future year’s cases given past years’ cases.

Results: We predicted the annual cases in 2016 and 2017 as 438 and 341, whereas the observed cases were 432 and 343, respectively. Pearson’s correlation coefficient and normalized root mean squared error values for 2016 and 2017 predictions were (0.83; 0.58) and (0.87; 0.52), respectively. The most important covariates were found to be the number of settlements with fewer than 25,000 inhabitants, latitude, longitude, and potential evapotranspiration (evaporation and transpiration).

Conclusions: Main driving factors of CCHF dynamics were human population at risk in rural areas, geographical dependency, and climate effect on ticks. Our model was able to prospectively predict the numbers of CCHF cases. Our proof-of-concept study also provided insight for understanding possible mechanisms of infectious diseases and found the important directions for practice and policy to combat against emerging infectious diseases.

Motivation: Survival analysis methods that integrate pathways/gene sets into their learning model could identify molecular mechanisms that determine survival characteristics of patients. Rather than first picking the predictive pathways/gene sets from a given collection and then training a predictive model on the subset of genomic features mapped to these selected pathways/gene sets, we developed a novel machine learning algorithm (Path2Surv) that conjointly performs these two steps using multiple kernel learning.

Results: We extensively tested our Path2Surv algorithm on 7655 patients from 20 cancer types using cancer-specific pathway/gene set collections and gene expression profiles of these patients. Path2Surv statistically significantly outperformed survival random forest (RF) on 12 out of 20 datasets and obtained comparable predictive performance against survival support vector machine (SVM) using significantly fewer gene expression features (i.e. less than 10% of what survival RF and survival SVM used).

Glioblastoma Multiforme (GBM) is the most common and aggressive primary brain tumor. Despite recent developments in surgery, chemo- and radio-therapy, a currently poor prognosis of GBM patients highlights an urgent need for novel treatment strategies. TRAIL (TNF Related Apoptosis Inducing Ligand) is a potent anti-cancer agent that can induce apoptosis selectively in cancer cells. GBM cells frequently develop resistance to TRAIL which renders clinical application of TRAIL therapeutics inefficient. In this study, we undertook a chemical screening approach using a library of epigenetic modifier drugs to identify compounds that could augment TRAIL response. We identified the fungal metabolite chaetocin, an inhibitor of histone methyl transferase SUV39H1, as a novel TRAIL sensitizer. Combining low subtoxic doses of chaetocin and TRAIL resulted in very potent and rapid apoptosis of GBM cells. Chaetocin also effectively sensitized GBM cells to further pro-apoptotic agents, such as FasL and BH3 mimetics. Chaetocin mediated apoptosis sensitization was achieved through ROS generation and consequent DNA damage induction that involved P53 activity. Chaetocin induced transcriptomic changes showed induction of antioxidant defense mechanisms and DNA damage response pathways. Heme Oxygenase 1 (HMOX1) was among the top upregulated genes, whose induction was ROS-dependent and HMOX1 depletion enhanced chaetocin mediated TRAIL sensitization. Finally, chaetocin and TRAIL combination treatment revealed efficacy in vivo. Taken together, our results provide a novel role for chaetocin as an apoptosis priming agent and its combination with pro-apoptotic therapies might offer new therapeutic approaches for GBMs.

Abstract: High mortality rates of glioblastoma (GBM) patients are partly attributed to the invasive behavior of tumor cells that exhibit extensive infiltration into adjacent brain tissue, leading to rapid, inevitable, and therapy-resistant recurrence. In this study, we analyzed transcriptome of motile (dispersive) and non-motile (core) GBM cells using an in vitro spheroid dispersal model and identified SERPINE1 as a modulator of GBM cell dispersal. Genetic or pharmacological inhibition of SERPINE1 reduced spheroid dispersal and cell adhesion by regulating cell-substrate adhesion. We examined TGFβ as a potential upstream regulator of SERPINE1 expression. We also assessed the significance of SERPINE1 in GBM growth and invasion using TCGA glioma datasets and a patient-derived orthotopic GBM model. SERPINE1 expression was associated with poor prognosis and mesenchymal GBM in patients. SERPINE1 knock-down in primary GBM cells suppressed tumor growth and invasiveness in the brain. Together, our results indicate that SERPINE1 is a key player in GBM dispersal and provide insights for future anti-invasive therapy design.

Keywords: GBM; transcriptome analysis; dispersal

Objectives: We aimed to describe the mechanisms of colistin resistance in Acinetobacter baumannii.

Methods: Twenty-nine patients diagnosed with colistin resistant A. baumannii infection were included to the study. The mutations in pmrCAB, lpxA, lpxC, and lpxD genes, expressions of pmrCAB, carbapenemases and mcr-1 positivity were studied.

Results: Twenty-seven (93%) of the patients received IV colistin therapy during their stay, and the case fatality rate was 45%. All of the mutations in pmrC and pmrB were found to be accompanied with a mutation in lpxD. The most common mutations were I42V and L150F in pmrC (65%), E117K in lpxD (65%), A138T in pmrB (58.6%). The colistin minimum inhibitory concentrations (MICs) of the isolates having any of these four mutations were higher than the isolates with no mutations (p < 0.001). The two most common mutations in pmrC (I42V and L150F) were found to be associated with higher expressions of pmrA and pmrC and higher colistin MIC values (p = 0.010 and 0.031). All isolates were bla_OXA-23 positive.

Conclusions: Coexistence of the lpxD mutation along with mutations in pmrCAB indicates synergistic function of these genes in development of colistin resistance in A. baumannii.

Missense mutations have various effects on protein structures, also leading to distorted protein dynamics that plausibly affects the function. We hypothesized that missense mutations in cancer-related genes selectively target hinge-neighboring residues that orchestrate collective structural dynamics. To test our hypothesis, we selected 69 cancer-related genes from the Cancer Gene Census database and their representative protein structures from the Protein Data Bank. We first identified the hinge residues in two global modes of motion by applying the Gaussian Network Model. We then showed that missense mutations are significantly enriched on hinge-neighboring residues in oncogenes and tumor suppressor genes. We observed that several oncogenes (eg, MAP2K1, PTPN11, and KRAS) and tumor suppressor genes (eg, EZH2, CDKN2C, and RHOA) strongly exhibit this phenomenon. This study highlights and rationalizes the functional importance of missense mutations on hinge-neighboring residues in cancer.

Background: Short and long ambulatory electrocardiographic monitoring with different systems is a widely used method to detect cardiac arrhythmias. In this study, we aimed to evaluate the effectiveness of a novel monitoring device on cardiac arrhythmia detection.

Methods: We used two different protocols to evaluate device performance. For the first one, 36 healthy subjects were enrolled. The standard 12-lead, 24-hour Holter monitoring and the novel single lead electrocardiogram (ECG) Patch Monitor (EPM) device (BeyondCare®, Rooti Labs Ltd., Taipei City, Taiwan) were simultaneously applied to all subjects for 24 hours. The quality of ECG data acquisition of novel system was compared to that of standard Holter. The second phase included 73 patients that were referred from our outpatient arrhythmia clinic for evaluation of their symptoms relevant to the cardiac arrhythmias. Advanced algorithms, statistical methods (cross-correlation method, Pearson’s correlation coefficient, Bland-Altman plots) were used to process and verify the acquired data.

Results: The overall average beat per minute correlation between BeyondCare® and standard 12-lead Holter was found 98% in 33 healthy subjects. The mean percentage of invalid measurements in BeyondCare® was 1.6% while the Holter’s was 1.7%. In the second protocol of the study, prospective data from 67 patients who were referred for evaluation of their symptoms relevant to cardiac arrhythmias, showed that the mean BeyondCare® wear time was 4.7±0.5 days out of five total days per protocol. The mean analyzable wear time was 93.6%. The water-resistant design enabled 73.5% of the participants to take a shower. 7.3% of participants had minor skin irritations related to the electrodes. Among the patients with detected arrhythmia (40.2% of all patients), 29.6% had their first arrhythmia after the initial two days period. A clinically significant pause was detected in one patient, ventricular tachycardia was detected in four patients, and supraventricular tachycardia was detected in 15 patients. Paroxysmal atrial fibrillation was identified in seven patients. Three of them had their first episodes after the second day of monitoring.

Conclusion: BeyondCare® Patch was well-tolerated and allowed prolonged time periods for continuous ECG monitoring, may result in an improvement in clinical accuracy and detection of arrhythmias by cloud-based artificial intelligence operating system.

Keywords: Ambulatory electrocardiographic monitoring, ECG patch monitoring device, Cardiac arrhythmias

Escherichia coli ST131 is a cause for global concern because of its high multidrug resistance and several virulence factors. In this study, the contribution of acrAB–TolC efflux system of E. coli ST131 to fluoroquinolone resistance was evaluated. A total of nonrepetitive 111 ciprofloxacin-resistant E. coli isolates were included in the study. Multilocus sequence typing was used for genotyping. Expressions of acrA, acrB, and TolC efflux pump genes were measured by RT-PCR. Mutations in marA, gyrA, parC, and aac(6′)-lb-cr positivity were studied by Sanger sequencing. Sixty-four (57.7%) of the isolates were classified as ST131, and 52 (81.3%) of the ST131 isolates belonged to H30-Rx subclone. In ST131, CTX-M 15 positivity (73%) and aac(6′)-lb-cr carriage (75%) were significantly higher than those in non-ST131 (12.8% and 51%, respectively) (P < 0.05). The ampicillin–sulbactam (83%) resistance was higher, and gentamicin resistance (20%) was lower in ST131 than that in non-ST131 (64% and 55%, respectively) (P = 0.001 and P = 0.0002). Numbers of the isolates with MDR or XDR profiles did not differ in both groups. Multiple in-dels (up to 16) were recorded in all quinolone-resistant isolates. However, marA gene was more overexpressed in ST131 compared to that in non-ST131 (median 5.98 vs. 3.99; P = 0.0007). Belonging to H30-Rx subclone, isolation site, ciprofloxacin MIC values did not correlate with efflux pump expressions. In conclusion, the marA regulatory gene of AcrAB–TolC efflux pump system has a significant impact on quinolone resistance and progression to MDR profile in ST131 clone. Efflux pump inhibitors might be alternative drugs for the treatment of infections caused by E. coli ST131 if used synergistically in combination with antibiotics.

We performed a systematic review and meta-analysis on the effectiveness of ribavirin use for the prevention of infection and death of healthcare workers exposed to patients with Crimean-Congo hemorrhagic fever virus (CCHFV) infection. Splashes with blood or bodily fluids (odds ratio [OR] 4.2), being a nurse or physician (OR 2.1), and treating patients who died from CCHFV infection (OR 3.8) were associated with healthcare workers acquiring CCHFV infection; 7% of the workers who received postexposure prophylaxis (PEP) with ribavirin and 89% of those who did not became infected. PEP with ribavirin reduced the odds of infection (OR 0.01, 95% CI 0–0.03), and ribavirin use ≤48 hours after symptom onset reduced the odds of death (OR 0.03, 95% CI 0–0.58). The odds of death increased 2.4-fold every day without ribavirin treatment. Ribavirin should be recommended as PEP and early treatment for workers at medium-to-high risk for CCHFV infection.

Background: Infectious diseases are one of the primary healthcare problems worldwide, leading to millions of deaths annually. To develop effective control and prevention strategies, we need reliable computational tools to understand disease dynamics and to predict future cases. These computational tools can be used by policy makers to make more informed decisions.

Methodology/Principal findings: In this study, we developed a computational framework based on Gaussian processes to perform spatiotemporal prediction of infectious diseases and exploited the special structure of similarity matrices in our formulation to obtain a very efficient implementation. We then tested our framework on the problem of modeling Crimean–Congo hemorrhagic fever cases between years 2004 and 2015 in Turkey.

Conclusions/Significance: We showed that our Gaussian process formulation obtained better results than two frequently used standard machine learning algorithms (i.e., random forests and boosted regression trees) under temporal, spatial, and spatiotemporal prediction scenarios. These results showed that our framework has the potential to make an important contribution to public health policy makers.

Motivation: Identifying molecular mechanisms that drive cancers from early to late stages is highly important to develop new preventive and therapeutic strategies. Standard machine learning algorithms could be used to discriminate early- and late-stage cancers from each other using their genomic characterizations. Even though these algorithms would get satisfactory predictive performance, their knowledge extraction capability would be quite restricted due to highly correlated nature of genomic data. That is why we need algorithms that can also extract relevant information about these biological mechanisms using our prior knowledge about pathways/gene sets.

Results: In this study, we addressed the problem of separating early- and late-stage cancers from each other using their gene expression profiles. We proposed to use a multiple kernel learning (MKL) formulation that makes use of pathways/gene sets (i) to obtain satisfactory/improved predictive performance and (ii) to identify biological mechanisms that might have an effect in cancer progression. We extensively compared our proposed MKL on gene sets algorithm against two standard machine learning algorithms, namely, random forests and support vector machines, on 20 diseases from the Cancer Genome Atlas cohorts for two different sets of experiments. Our method obtained statistically significantly better or comparable predictive performance on most of the datasets using significantly fewer gene expression features. We also showed that our algorithm was able to extract meaningful and disease-specific information that gives clues about the progression mechanism.

The Fbw7 (F-box/WD repeat-containing protein 7) ubiquitin ligase targets multiple oncoproteins for degradation and is commonly mutated in cancers. Like other pleiotropic tumor suppressors, Fbw7’s complex biology has impeded our understanding of how Fbw7 mutations promote tumorigenesis and hindered the development of targeted therapies. To address these needs, we employed a transfer learning approach to derive gene-expression signatures from The Cancer Gene Atlas datasets that predict Fbw7 mutational status across tumor types and identified the pathways enriched within these signatures. Genes involved in mitochondrial function were highly enriched in pan-cancer signatures that predict Fbw7 mutations. Studies in isogenic colorectal cancer cell lines that differed in Fbw7 mutational status confirmed that Fbw7 mutations increase mitochondrial gene expression. Surprisingly, Fbw7 mutations shifted cellular metabolism toward oxidative phosphorylation and caused context-specific metabolic vulnerabilities. Our approach revealed unexpected metabolic reprogramming and possible therapeutic targets in Fbw7-mutant cancers and provides a framework to study other complex, oncogenic mutations.

Objectives: We describe molecular characteristics of colistin resistance and its impact on patient mortality.

Methods: A prospective cohort study was performed in seven different Turkish hospitals. The genotype of each isolate was determined by MLST and repetitive extragenic palindromic PCR (rep-PCR). Alterations in the mgrB were detected by sequencing. Upregulation of pmrCAB, phoQ and pmrK was quantified by RT-PCR. mcr-1 and the genes encoding OXA-48, NDM-1 and KPC were amplified by PCR.

Results: A total of 115 patients diagnosed with colistin-resistant K. pneumoniae (ColR-Kp) infection were included. Patients were predominantly males (55%) with a median age of 63 (IQR 46–74) and the 30 day mortality rate was 61%. ST101 was the most common ST and accounted for 68 (59%) of the ColR-Kp. The 30 day mortality rate in patients with these isolates was 72%. In ST101, 94% (64/68) of the isolates had an altered mgrB gene, whereas the alteration occurred in 40% (19/47) of non-ST101 isolates. The OXA-48 and NDM-1 carbapenemases were found in 93 (81%) and 22 (19%) of the total 115 isolates, respectively. In multivariate analysis for the prediction of 30 day mortality, ST101 (OR 3.4, CI 1.46–8.15, P = 0.005) and ICU stay (OR 7.4, CI 2.23–29.61, P = 0.002) were found to be significantly associated covariates.

Conclusions: Besides ICU stay, ST101 was found to be a significant independent predictor of patient mortality among those infected with ColR-Kp. A significant association was detected between ST101 and OXA-48. ST101 may become a global threat in dissemination of colistin resistance and increased morbidity and mortality of K.pneumoniae infection.

We report the results of a DREAM challenge designed to predict preferential/relative genetic vulnerabilities/essentialities based on a novel data set testing 98,000 shRNAs against 149 cancer cell lines. We analyzed the results of over 3,000 submissions over a period of four months. We found that algorithms combining essentiality data across multiple genes demonstrated increased accuracy; gene expression was the most informative molecular data; the identity of the gene being predicted was far more important than the modeling strategy; well-predicted genes and selected molecular features showed enrichment in functional categories; and frequently selected expression features correlate with survival in primary tumors. This study establishes benchmarks for gene essentiality prediction, presents a community resource for future comparison against this benchmark, and provides insights into factors influencing the ability to predict gene essentiality from functional genetic screens. This study demonstrated the value of releasing pre-publication data publicly to engage the community in an open research collaboration.

Background: We described the predictive role of cytokines in fatality of Crimean-Congo Hemorrhagic Fever Virus (CCHFV) infection by using daily clinical sera samples.

Methods: Consequent serum samples of the selected patients in different severity groups and healthy controls were examined by using human cytokine 17-plex assay.

Results: We included 12 (23%) mild, 30 (58%) moderate, 10 (19%) severe patients, and 10 healthy volunteers. The mean age of the patients was 52 (sd 15), 52% were female. Forty-six patients (88%) received ribavirin. During disease course, the median levels of IL-6, IL-8, IL-10, IL-10/12, IFN-γ, MCP-1 and MIP-1b were found to be significantly higher among CCHF patients than the healthy controls. Within the first five days after onset of disease, among the fatal cases, the median levels of IL-6 and IL-8 were found to be significantly higher than the survived ones (Figure 3), and MCP-1 was elevated among fatal cases, but statistical significance was not detected.In receiver operating characteristic (ROC) analysis, IL-8 (92%), IL-6 (92%), MCP-1 (79%) were found to be the most significant cytokines in predicting the fatality rates in the early period of the disease (5 days).

Conclusion: IL-6 and IL-8 can predict the poor outcome, within the first five days of disease course. Elevated IL-6 and IL-8 levels within first five days could be used as prognostic markers.

Motivation: In recent years, vast advances in biomedical technologies and comprehensive sequencing have revealed the genomic landscape of common forms of human cancer in unprecedented detail. The broad heterogeneity of the disease calls for rapid development of personalized therapies. Translating the readily available genomic data into useful knowledge that can be applied in the clinic remains a challenge. Computational methods are needed to aid these efforts by robustly analyzing genome-scale data from distinct experimental platforms for prioritization of targets and treatments.

Results: We propose a novel, biologically-motivated, Bayesian multitask approach, which explicitly models gene-centric dependencies across multiple and distinct genomic platforms. We introduce a genewise prior and present a fully Bayesian formulation of a group factor analysis model. In supervised prediction applications, our multitask approach leverages similarities in response profiles of groups of drugs that are more likely to be related to true biological signal, which leads to more robust performance and improved generalization ability. We evaluate the performance of our method on molecularly characterized collections of cell lines profiled against two compound panels, namely the Cancer Cell Line Encyclopedia and the Cancer Therapeutics Response Portal. We demonstrate that accounting for the gene-centric dependencies enables leveraging information from multi-omic input data and improves prediction and feature selection performance. We further demonstrate the applicability of our method in an unsupervised dimensionality reduction application by inferring genes essential to tumorigenesis in the pancreatic ductal adenocarcinoma and lung adenocarcinoma patient cohorts from The Cancer Genome Atlas.

Motivation: Identifying molecular signatures of disease phenotypes is studied using two mainstream approaches: (i) Predictive modeling methods such as linear classification and regression algorithms are used to find signatures predictive of phenotypes from genomic data, which may not be robust due to limited sample size or highly correlated nature of genomic data. (ii) Gene set analysis methods are used to find gene sets on which phenotypes are linearly dependent by bringing prior biological knowledge into the analysis, which may not capture more complex nonlinear dependencies. Thus, formulating an integrated model of gene set analysis and nonlinear predictive modeling is of great practical importance.

Results: In this study, we propose a Bayesian binary classification framework to integrate gene set analysis and nonlinear predictive modeling. We then generalize this formulation to multitask learning setting to model multiple related datasets conjointly. Our main novelty is the probabilistic nonlinear formulation that enables us to robustly capture nonlinear dependencies between genomic data and phenotype even with small sample sizes. We demonstrate the performance of our algorithms using repeated random subsampling validation experiments on two cancer and two tuberculosis datasets by predicting important disease phenotypes from genome-wide gene expression data. We are able to obtain comparable or even better predictive performance than a baseline Bayesian nonlinear algorithm and to identify sparse sets of relevant genes and gene sets on all datasets. We also show that our multitask learning formulation enables us to further improve the generalization performance and to better understand biological processes behind disease phenotypes.

Many important signaling and regulatory proteins are expressed at low abundance and are difficult to measure in single cells. We report a molecular imaging approach to quantitate protein levels by digitized, discrete counting of nanoparticle-tagged proteins. Digitized protein counting provides ultrasensitive molecular detection of proteins in single cells that surpasses conventional methods of quantitating total diffuse fluorescence, and offers a substantial improvement in protein quantitation. We implement this digitized proteomic approach in an integrated imaging platform, the single cell-quantum dot platform (SC-QDP), to execute sensitive single cell phosphoquantitation in response to multiple drug treatment conditions and using limited primary patient material. The SC-QDP: 1) identified pAKT and pERK phospho-heterogeneity and insensitivity in individual leukemia cells treated with a multi-drug panel of FDA-approved kinase inhibitors, and 2) revealed subpopulations of drug-insensitive CD34+ stem cells with high pCRKL and pSTAT5 signaling in chronic myeloid leukemia patient blood samples. This ultrasensitive digitized protein detection approach is valuable for uncovering subtle but important differences in signaling, drug insensitivity, and other key cellular processes amongst single cells.

Affective classification and retrieval of multimedia such as audio, image, and video have become emerging research areas in recent years. The previous research focused on designing features and developing feature extraction methods. Generally, a multimedia content can be represented with different feature representations (i.e., views). However, the most suitable feature representation related to people׳s emotions is usually not known a priori. We propose here a novel Bayesian multiple kernel learning algorithm for affective classification and retrieval tasks. The proposed method can make use of different representations simultaneously (i.e., multiview learning) to obtain a better prediction performance than using a single feature representation (i.e., single-view learning) or a subset of features, with the advantage of automatic feature selections. In particular, our algorithm has been implemented within a multilabel setup to capture the correlation between emotions, and the Bayesian formulation enables our method to produce probabilistic outputs for measuring a set of emotions triggered by a single image. As a case study, we perform classification and retrieval experiments with our algorithm for predicting people׳s emotional states evoked by images, using generic low-level image features. The empirical results with our approach on the widely-used International Affective Picture System (IAPS) data set outperform several existing methods in terms of classification performance and results interpretability.

Predicting the best treatment strategy from genomic information is a core goal of precision medicine. Here we focus on predicting drug response based on a cohort of genomic, epigenomic and proteomic profiling data sets measured in human breast cancer cell lines. Through a collaborative effort between the National Cancer Institute (NCI) and the Dialogue on Reverse Engineering Assessment and Methods (DREAM) project, we analyzed a total of 44 drug sensitivity prediction algorithms. The top-performing approaches modeled nonlinear relationships and incorporated biological pathway information. We found that gene expression microarrays consistently provided the best predictive power of the individual profiling data sets; however, performance was increased by including multiple, independent data sets. We discuss the innovations underlying the top-performing methodology, Bayesian multitask MKL, and we provide detailed descriptions of all methods. This study establishes benchmarks for drug sensitivity prediction and identifies approaches that can be leveraged for the development of new methods.

We extend kernelized matrix factorization with a full-Bayesian treatment and with an ability to work with multiple side information sources expressed as different kernels. Kernels have been introduced to integrate side information about the rows and columns, which is necessary for making out-of-matrix predictions. We discuss specifically binary output matrices but extensions to realvalued matrices are straightforward. We extend the state of the art in two key aspects: (i) A full-conjugate probabilistic formulation of the kernelized matrix factorization enables an efficient variational approximation, whereas full-Bayesian treatments are not computationally feasible in the earlier approaches. (ii) Multiple side information sources are included, treated as different kernels in multiple kernel learning which additionally reveals which side sources are informative. We then show that the framework can also be used for supervised and semi-supervised multilabel classification and multi-output regression, by considering samples and outputs as the domains where matrix factorization operates. Our method outperforms alternatives in predicting drug-protein interactions on two data sets. On multilabel classification, our algorithm obtains the lowest Hamming losses on 10 out of 14 data sets compared to five state-of-the-art multilabel classification algorithms. We finally show that the proposed approach outperforms alternatives in multi-output regression experiments on a yeast cell cycle data set.

Motivation: Human immunodeficiency virus (HIV) and cancer require personalized therapies owing to their inherent heterogeneous nature. For both diseases, large-scale pharmacogenomic screens of molecularly characterized samples have been generated with the hope of identifying genetic predictors of drug susceptibility. Thus, computational algorithms capable of inferring robust predictors of drug responses from genomic information are of great practical importance. Most of the existing computational studies that consider drug susceptibility prediction against a panel of drugs formulate a separate learning problem for each drug, which cannot make use of commonalities between subsets of drugs.

Results: In this study, we propose to solve the problem of drug susceptibility prediction against a panel of drugs in a multitask learning framework by formulating a novel Bayesian algorithm that combines kernel-based non-linear dimensionality reduction and binary classification (or regression). The main novelty of our method is the joint Bayesian formulation of projecting data points into a shared subspace and learning predictive models for all drugs in this subspace, which helps us to eliminate off-target effects and drug-specific experimental noise. Another novelty of our method is the ability of handling missing phenotype values owing to experimental conditions and quality control reasons. We demonstrate the performance of our algorithm via cross-validation experiments on two benchmark drug susceptibility datasets of HIV and cancer. Our method obtains statistically significantly better predictive performance on most of the drugs compared with baseline single-task algorithms that learn drug-specific models. These results show that predicting drug susceptibility against a panel of drugs simultaneously within a multitask learning framework improves overall predictive performance over single-task learning approaches.

With data from recent large-scale drug sensitivity measurement campaigns, it is now possible to build and test models predicting responses for more than one hundred anticancer drugs against several hundreds of human cancer cell lines. Traditional quantitative structure–activity relationship (QSAR) approaches focus on small molecules in searching for their structural properties predictive of the biological activity in a single cell line or a single tissue type. We extend this line of research in two directions: (1) an integrative QSAR approach predicting the responses to new drugs for a panel of multiple known cancer cell lines simultaneously and (2) a personalized QSAR approach predicting the responses to new drugs for new cancer cell lines. To solve the modeling task, we apply a novel kernelized Bayesian matrix factorization method. For maximum applicability and predictive performance, the method optionally utilizes genomic features of cell lines and target information on drugs in addition to chemical drug descriptors. In a case study with 116 anticancer drugs and 650 cell lines, we demonstrate the usefulness of the method in several relevant prediction scenarios, differing in the amount of available information, and analyze the importance of various types of drug features for the response prediction. Furthermore, after predicting the missing values of the data set, a complete global map of drug response is explored to assess treatment potential and treatment range of therapeutically interesting anticancer drugs.

Several computational approaches have been proposed for inferring the affective state of the user, motivated for example by the goal of building improved interfaces that can adapt to the user׳s needs and internal state. While fairly good results have been obtained for inferring the user state under highly controlled conditions, a considerable amount of work remains to be done for learning high-quality estimates of subjective evaluations of the state in more natural conditions. In this work, we discuss how two recent machine learning concepts, multi-view learning and multi-task learning, can be adapted for user state recognition, and demonstrate them on two data collections of varying quality. Multi-view learning enables combining multiple measurement sensors in a justified way while automatically learning the importance of each sensor. Multi-task learning, in turn, tells how multiple learning tasks can be learned together to improve the accuracy. We demonstrate the use of two types of multi-task learning: learning both multiple state indicators and models for multiple users together. We also illustrate how the benefits of multi-task learning and multi-view learning can be effectively combined in a unified model by introducing a novel algorithm.

Coupled training of dimensionality reduction and classification is proposed previously to improve the prediction performance for single-label problems. Following this line of research, in this paper, we first introduce a novel Bayesian method that combines linear dimensionality reduction with linear binary classification for supervised multilabel learning and present a deterministic variational approximation algorithm to learn the proposed probabilistic model. We then extend the proposed method to find intrinsic dimensionality of the projected subspace using automatic relevance determination and to handle semi-supervised learning using a low-density assumption. We perform supervised learning experiments on four benchmark multilabel learning data sets by comparing our method with baseline linear dimensionality reduction algorithms. These experiments show that the proposed approach achieves good performance values in terms of hamming loss, average AUC, macro F1, and micro F1 on held-out test data. The low-dimensional embeddings obtained by our method are also very useful for exploratory data analysis. We also show the effectiveness of our approach in finding intrinsic subspace dimensionality and semi-supervised learning tasks.

Dimensionality reduction is commonly used as a preprocessing step before training a supervised learner. However, coupled training of dimensionality reduction and supervised learning steps may improve the prediction performance. In this paper, we introduce a simple and novel Bayesian supervised dimensionality reduction method that combines linear dimensionality reduction and linear supervised learning in a principled way. We present both Gibbs sampling and variational approximation approaches to learn the proposed probabilistic model for multiclass classification. We also extend our formulation towards model selection using automatic relevance determination in order to find the intrinsic dimensionality. Classification experiments on three benchmark data sets show that the new model significantly outperforms seven baseline linear dimensionality reduction algorithms on very low dimensions in terms of generalization performance on test data. The proposed model also obtains the best results on an image recognition task in terms of classification and retrieval performances.

For supervised learning problems, dimensionality reduction is generally applied as a preprocessing step. However, coupled training of dimensionality reduction and supervised learning steps may improve the prediction performance. In this paper, we propose a novel dimensionality reduction algorithm coupled with a supervised kernel-based learner, called supervised multiple kernel embedding, that integrates multiple kernel learning to dimensionality reduction and performs prediction on the projected subspace with a joint optimization framework. Combining multiple kernels allows us to combine different feature representations and/or similarity measures toward a unified subspace. We perform experiments on one digit recognition and two bioinformatics data sets. Our proposed method significantly outperforms multiple kernel Fisher discriminant analysis followed by a standard kernel-based learner, especially on low dimensions.

Instead of selecting a single kernel, multiple kernel learning (MKL) uses a weighted sum of kernels where the weight of each kernel is optimized during training. Such methods assign the same weight to a kernel over the whole input space, and we discuss localized multiple kernel learning (LMKL) that is composed of a kernel-based learning algorithm and a parametric gating model to assign local weights to kernel functions. These two components are trained in a coupled manner using a two-step alternating optimization algorithm. Empirical results on benchmark classification and regression data sets validate the applicability of our approach. We see that LMKL achieves higher accuracy compared with canonical MKL on classification problems with different feature representations. LMKL can also identify the relevant parts of images using the gating model as a saliency detector in image recognition problems. In regression tasks, LMKL improves the performance significantly or reduces the model complexity by storing significantly fewer support vectors.

Motivation: Identifying interactions between drug compounds and target proteins has a great practical importance in the drug discovery process for known diseases. Existing databases contain very few experimentally validated drug–target interactions and formulating successful computational methods for predicting interactions remains challenging.

Results: In this study, we consider four different drug–target interaction networks from humans involving enzymes, ion channels, G-protein-coupled receptors and nuclear receptors. We then propose a novel Bayesian formulation that combines dimensionality reduction, matrix factorization and binary classification for predicting drug–target interaction networks using only chemical similarity between drug compounds and genomic similarity between target proteins. The novelty of our approach comes from the joint Bayesian formulation of projecting drug compounds and target proteins into a unified subspace using the similarities and estimating the interaction network in that subspace. We propose using a variational approximation in order to obtain an efficient inference scheme and give its detailed derivations. Lastly, we demonstrate the performance of our proposed method in three different scenarios: (i) exploratory data analysis using low-dimensional projections, (ii) predicting interactions for the out-of-sample drug compounds and (iii) predicting unknown interactions of the given network.

Many financial organizations such as banks and retailers use computational credit risk analysis (CRA) tools heavily due to recent financial crises and more strict regulations. This strategy enables them to manage their financial and operational risks within the pool of financial institutes. Machine learning algorithms especially binary classifiers are very popular for that purpose. In real-life applications such as CRA, feature selection algorithms are used to decrease data acquisition cost and to increase interpretability of the decision process. Using feature selection methods directly on CRA data sets may not help due to categorical variables such as marital status. Such features are usually are converted into binary features using 1-of-k encoding and eliminating a subset of features from a group does not help in terms of data collection cost or interpretability. In this study, we propose to use the probit classifier with a proper prior structure and multiple kernel learning with a proper kernel construction procedure to perform group-wise feature selection (i.e., eliminating a group of features together if they are not helpful). Experiments on two standard CRA data sets show the validity and effectiveness of the proposed binary classification algorithm variants.

In recent years, several methods have been proposed to combine multiple kernels instead of using a single one. These different kernels may correspond to using different notions of similarity or may be using information coming from multiple sources (different representations or different feature subsets). In trying to organize and highlight the similarities and differences between them, we give a taxonomy of and review several multiple kernel learning algorithms. We perform experiments on real data sets for better illustration and comparison of existing algorithms. We see that though there may not be large differences in terms of accuracy, there is difference between them in complexity as given by the number of stored support vectors, the sparsity of the solution as given by the number of used kernels, and training time complexity. We see that overall, using multiple kernels instead of a single one is useful and believe that combining kernels in a nonlinear or data-dependent way seems more promising than linear combination in fusing information provided by simple linear kernels, whereas linear methods are more reasonable when combining complex Gaussian kernels.

In recent years, several methods have been proposed to combine multiple kernels using a weighted linear sum of kernels. These different kernels may be using information coming from multiple sources or may correspond to using different notions of similarity on the same source. We note that such methods, in addition to the usual ones of the canonical support vector machine formulation, introduce new regularization parameters that affect the solution quality and, in this work, we propose to optimize them using response surface methodology on cross-validation data. On several bioinformatics and digit recognition benchmark data sets, we compare multiple kernel learning and our proposed regularized variant in terms of accuracy, support vector count, and the number of kernels selected. We see that our proposed variant achieves statistically similar or higher accuracy results by using fewer kernel functions and/or support vectors through suitable regularization; it also allows better knowledge extraction because unnecessary kernels are pruned and the favored kernels reflect the properties of the problem at hand.

We formulate a supervised, localized dimensionality reduction method using a gating model that divides up the input space into regions and selects the dimensionality reduction projection separately in each region. The gating model, the locally linear projections, and the kernel-based supervised learning algorithm which uses them in its kernels are coupled and their training is performed with an alternating optimization procedure. Our proposed local projection kernel projects a data instance into different feature spaces by using the local projection matrices, combines them with the gating model, and performs the dot product in the combined feature space. Empirical results on benchmark data sets for visualization and classification tasks validate the idea. The method is generalizable to regression estimation and novelty detection.

Recently, it has been proposed to combine multiple kernels using a weighted linear sum. In certain applications, different kernels may be using different input representations and these methods do not consider neither the cost of acquiring them nor the cost of evaluating the kernels. We generalize the framework of Multiple Kernel Learning (MKL) for this cost-conscious methodology. On 12 benchmark data sets from the UCI repository, we compare MKL and its cost-conscious variants in terms of accuracy, support vector count, and total cost. Cost-conscious MKL achieves statistically similar accuracy results by using fewer support vectors/kernels by best trading off accuracy brought by each representation/kernel with the concomitant cost. We also test our approach on two popular bioinformatics data sets from MIPS comprehensive yeast genome database (CYGD) and see that integrating the cost factor into kernel combination allows us to obtain cheaper kernel combinations by using fewer active kernels and/or support vectors.

Background: Computational prediction of protein stability change due to single-site amino acid substitutions is of interest in protein design and analysis. We consider the following four ways to improve the performance of the currently available predictors: (1) We include additional sequence- and structure-based features, namely, the amino acid substitution likelihoods, the equilibrium fluctuations of the alpha- and beta-carbon atoms, and the packing density. (2) By implementing different machine learning integration approaches, we combine information from different features or representations. (3) We compare classification vs. regression methods to predict the sign vs. the output of stability change. (4) We allow a reject option for doubtful cases where the risk of misclassification is high.

Results: We investigate three different approaches: early, intermediate and late integration, which respectively combine features, kernels over feature subsets, and decisions. We perform simulations on two data sets: (1) S1615 is used in previous studies, (2) S2783 is the updated version (as of July 2, 2009) extracted also from ProTherm. For S1615 data set, our highest accuracy using both sequence and structure information is 0.842 on cross-validation and 0.904 on testing using early integration. Newly added features, namely, local compositional packing and the mobility extent of the mutated residues, improve accuracy significantly with intermediate integration. For S2783 data set, we also train regression methods to estimate not only the sign but also the amount of stability change and apply risk-based classification to reject when the learner has low confidence and the loss of misclassification is high. The highest accuracy is 0.835 on cross-validation and 0.832 on testing using only sequence information. The percentage of false positives can be decreased to less than 0.005 by rejecting 10 per cent using late integration.

Conclusions: We find that in both early and late integration, combining inputs or decisions is useful in increasing accuracy. Intermediate integration allows assessing the contributions of individual features by looking at the assigned weights. Overall accuracy of regression is not better than that of classification but it has less false positives, especially when combined with the reject option. The server for stability prediction for three integration approaches and the data sets are available at http://www.prc.boun.edu.tr/appserv/prc/mlsta.

Tao et al. have recently proposed the posterior probability support vector machine (PPSVM) which uses soft labels derived from estimated posterior probabilities to be more robust to noise and outliers. Tao et al.’s model uses a window-based density estimator to calculate the posterior probabilities and is a binary classifier. We propose a neighbor-based density estimator and also extend the model to the multiclass case. Our bias–variance analysis shows that the decrease in error by PPSVM is due to a decrease in bias. On 20 benchmark data sets, we observe that PPSVM obtains accuracy results that are higher or comparable to those of canonical SVM using significantly fewer support vectors.