ROC AUC Values Research Articles

Five long non-coding RNAs establish a prognostic nomogram and construct a competing endogenous RNA network in the progression of non-small cell lung cancer

BackgroundAccumulating evidence has revealed that long non-coding RNAs (lncRNAs) play vital roles in the progression of non-small cell lung cancer (NSCLC). But the relationship between lncRNAs and survival outcome of NSCLC remains to be explored. Therefore, we attempt to figure out their survival roles and molecular connection in NSCLC.MethodsBy analyzing the transcriptome profiling of NSCLC from TCGA databases, we divided patients into three groups, and identified differentially expressed lncRNAs (DELs) of each group. Next, we explored the prognostic roles of common DELs by univariate and multivariate Cox analysis, Lasson, and Kaplan-Meier analysis. Additionally, we assessed and compared the prognostic accuracy of 5 lncRNAs through ROC curves and AUC values. Ultimately, we detected their potential function by enrichment analysis and molecular connection through establishing a competing endogenous RNA (ceRNA) network.ResultsOne hundred ninety-seven common DELs were spotted. And we successfully screened out 5 lncRNAs related to the patient’s survival, including LINC01833, AC112206.2, FAM83A-AS1, BANCR, and HOTAIR. Combing with age and AJCC stage, we constructed a nomogram that prognostic prediction was superior to the traditional parameters. Furthermore, 275 qualified mRNAs related to 5 lncRNAs were spotted. Functional analysis indicates that these lncRNAs act key roles in the progression of NSCLC, such as P53 and cell cycle signaling pathway. And ceRNA network also suggests that these lncRNAs are tightly connected with tumor progression.ConclusionsA nomogram and ceRNA network based on 5 lncRNAs indicate that there can effectively predict the overall survival of NSCLC and potentially serve as a therapeutic guide for NSCLC.

Urine protein biomarkers of bladder cancer arising from 16-plex antibody-based screens.

Purpose: The purpose of this study is to identify novel urine protein biomarkers of bladder cancer using a Luminex based screening platform.Materials and Methods: The current study examines urine samples from 66 subjects, comprised of 31 Urology clinic controls and 35 bladder cancer patients, using a Luminex based screening platform. ELISA validation was carried out for the top 4 prospective urine biomarkers using an independent cohort of 20 Urology clinic controls and 60 bladder cancer (BC) subjects.Results: Of the 16 proteins screened by Luminex, 10 showed significant elevation in BC compared to the controls. Eight of these urine proteins were able to differentiate BC from control urine with ROC AUC values exceeding 0.70 at p < 0.0001, with specificity values exceeding 0.9. Upon ELISA validation, urine IL-1α, IL-1ra, and IL-8 were able to distinguish control urine from urine drawn from various bladder cancer stages, with IL-8 being the best discriminator. Compared to members of the IL-1 cytokine family, urine IL-8 was also best at discriminating T1 and/or T2–T4 from Ta BC (ROC AUC ≥ 0.83), as well as high grade from low grade BC (ROC AUC ≥ 0.82).Conclusions: These findings suggest that urine IL-1α, IL-1ra and IL-8 are useful indicators of bladder cancer. Urine IL-8 not only distinguishes bladder cancer from controls, it also discriminates high grade from low grade disease, and the successive clinical stages of bladder cancer. While supportive of previous reports, these findings warrant further analysis in prospective cohorts.

The atherosclerosis burden score.

Purpose: We carried out this study to evaluate the predictive value of atherosclerosis burden score (ABS) to predict coronary artery disease (CAD) among asymptomatic patients without known cardiovascular disease (CVD), as compared to other imaging or functional techniques, namely coronary artery calcium (CAC) score, carotid intima-media thickness (C-IMT), and ankle brachial index (ABI). Patients and methods: This prospective study included 198 asymptomatic consecutive patients referred for evaluation of their cardiovascular (CV) risk and for therapeutic advice. Traditional CV risk factors, ABS, CAC score, C-IMT, ABI and an ECG-synchronized coronary CT-angiography (CCTA) were performed for each patient. We compared the predictive values of these atherosclerosis markers to detect CAD defined as coronary stenosis ≥30% objectivated by CCTA. Results: Among the whole sample, the area under the receiver-operating characteristic curve (ROC-AUC) was significantly higher for CAC score (0.81, p=0.015) than for ABS, the reference (0.70) but these values were lower for C-IMT (0.60, p=0.16) and particularly for ABI (0.56, p=0.0015). However, among patients at intermediate risk of coronary heart disease (CHD), according to Framingham risk score (FRS), the differences between the ROC-AUC values for ABS (0.70) and CAC score (0.76, p=0.36) were less pronounced. Again, as compared to ABS, the ROC-AUC values were lower for C-IMT (0.60, p=0.21) and ABI (0.57, p=0.06). Conclusions: ABS, an ultrasonographic score based on the assessment of carotid and femoral plaque burden, predicts more accurately CAD than other non-radiation tools analyzed here, and has a similar performance to CAC in patients at intermediate CHD risk. Thus, ABS could be an appropriate non-invasive and safe method to improve the detection of high-risk patients who will benefit from a more intensive therapy for the primary prevention of CVD.

Abstract PO-054: Development of machine learning based early warning for rapid response system in a single cancer center

Abstract Background: Rapid response systems identify deteriorating hospitalized patients and prevent unexpected mortality by providing immediate interventions. But there are also difficulties in costs and human resources for clinical implementation. Machine learning is an application of artificial intelligence that provides systems the ability to automatically learn and improve from experience without being explicitly programmed. We try to develop a machine learning based early warning system to predict intubation and cardiopulmonary resuscitation (CPR) using vital signs and laboratory values. Methods: The training set was composed of all patients admitted to a cancer hospital named Chonnam National University Hwasun Hospital for two years (event group 277 vs. non-event group 3576 cases), and the test set was made from all patients admitted to non-cancer hospital named Chonnam National University Hospital for four months (event group 61 vs. non-event group 2334 cases). The time point at which the patient's clinical deterioration was recognized in the ward was defined as ‘detection time’, and the time point at which the patient suffered airway intubation or CPR was defined as ‘event time’. The algorithm is developed using five vital signs (blood pressure, heart rate, body temperature, respiratory rate and oxygen saturation) and laboratory values. We compared the efficacy of developed models which were ensemble model, recurrent neural network (RNN), low gradient boosting (LGB), extreme gradient boosting (XGB) and random forest with established modified early warning score (MEWS) using receiver operating characteristic area under curve (ROC AUC) and precision recall AUC (PR AUC) analyses. Results: The ROC AUC value of the ensemble model using five vital signs was 0.9219 and the PR AUC was 0.0732. After adding laboratory values, the ROC AUC values of the ensemble, RNN, LGB, XGB, random forest improved to 0.9964, 0.9814, 0.9950, 0.9862 and 0.9962. The PR AUC values were also increased to 0.2326, 0.2140, 0.3252, 0.2095 and 0.1985, respectively. The developed models showed higher performance than MEWS (ROC AUC: 0.7253, PR AUC: 0.0468) Conclusion: Machine learning based early warning system using vital signs and laboratory results was feasible. And the performance was better than traditional MEWS model because the ensemble model showed highest value in ROC AUC and LGB showed highest value in PR AUC. Citation Format: Bo-Gun Kho, Min-Seok Kim, Tae-Ok Kim, Cheol-Kyu Park, Young-Chul Kim, Soo-Hyung Kim, In-Jae Oh. Development of machine learning based early warning for rapid response system in a single cancer center [abstract]. In: Proceedings of the AACR Virtual Special Conference on Artificial Intelligence, Diagnosis, and Imaging; 2021 Jan 13-14. Philadelphia (PA): AACR; Clin Cancer Res 2021;27(5_Suppl):Abstract nr PO-054.

Fast Identification of Adverse Drug Reactions (ADRs) of Digestive and Nervous Systems of Organic Drugs by In Silico Models.

This study aimed to discover concurrences of adverse drug reactions (ADRs) and derive models of the most frequent items of ADRs based on the SIDER database, which included 1430 marketed drugs and 5868 ADRs. First, common ADRs of organic drugs were manually reclassified according to side effects in the human system and followed by an association rule analysis, which found ADRs of digestive and nervous systems often occurred at the same time with a good association rule. Then, three algorithms, linear discriminant analysis (LDA), support vector machine (SVM) and deep learning, were used to derive models of ADRs of digestive and nervous systems based on 497 organic monomer drugs and to identify key structural features in defining these ADRs. The statistical results indicated that these kinds of QSAR models were good tools for screening ADRs of digestive and nervous systems, which gave the ROC AUC values of 81.5%, 98.9%, 91.5%, 69.5%, 78.4% and 78.8%, respectively. Then, these models were applied to investigate ADRs of 1536 organic compounds with four phase and zero rule-of-five (RO5) violations from the ChEMBL database. Based on the consensus ADRs’ predictions of models, 58.1% and 42.6% of compounds were predicted to cause these two ADRs, respectively, indicating the significance of initial assessment of ADRs in early drug discovery.

Artificial intelligence for detection of periapical lesions on intraoral radiographs: Comparison between convolutional neural networks and human observers

The aim of this study was to compare the diagnostic performance of convolutional neural networks (CNNs) with the performance of human observers for the detection of simulated periapical lesions on periapical radiographs. Ten sockets were prepared in bovine ribs. Periapical defects of 3 sizes were sequentially created. Periapical radiographs were acquired of each socket with no lesion and with each lesion size with a photostimulable storage phosphor system. Radiographs were evaluated with no filter and with 6 image filter settings. A CNN architecture was set up using Keras-TensorFlow. Separate CNNs were evaluated for randomly sampled training/validation data and for data split up by socket (5-fold cross-validation) and filter (7-fold cross-validation). CNN performance on validation data was compared with that of 3 oral radiologists for sensitivity, specificity, and area under the receiver operating characteristic curve (ROC-AUC). Using random sampling, the CNN showed perfect accuracy for the validation data. When data were split up by socket, the mean sensitivity, specificity, and ROC-AUC values were 0.79, 0.88, and 0.86, respectively; when split up by filter, they were 0.87, 0.98, and 0.93, respectively. For radiologists, the values were 0.58, 0.83, and 0.75, respectively. CNNs show promise in periapical lesion detection. The pretrained CNN model yielded in this study can be used for further training on larger samples and/or clinical radiographs.

WIDOCK: a reactive docking protocol for virtual screening of covalent inhibitors

Here we present WIDOCK, a virtual screening protocol that supports the selection of diverse electrophiles as covalent inhibitors by incorporating ligand reactivity towards cysteine residues into AutoDock4. WIDOCK applies the reactive docking method (Backus et al. in Nature 534:570–574, 2016) and extends it into a virtual screening tool by introducing facile experimental or computational parametrization and a ligand focused evaluation scheme together with a retrospective and prospective validation against various therapeutically relevant targets. Parameters accounting for ligand reactivity are derived from experimental reaction kinetic data or alternatively from computed reaction barriers. The performance of this docking protocol was first evaluated by investigating compound series with diverse warhead chemotypes against KRASG12C, MurA and cathepsin B. In addition, WIDOCK was challenged on larger electrophilic libraries screened against OTUB2 and NUDT7. These retrospective analyses showed high sensitivity in retrieving experimental actives, by also leading to superior ROC curves, AUC values and better enrichments than the standard covalent docking tool available in AutoDock4 when compound collections with diverse warheads were investigated. Finally, we applied WIDOCK for the prospective identification of covalent human MAO-A inhibitors acting via a new mechanism by binding to Cys323. The inhibitory activity of several predicted compounds was experimentally confirmed and the labelling of Cys323 was proved by subsequent MS/MS measurements. These findings demonstrate the usefulness of WIDOCK as a warhead-sensitive, covalent virtual screening protocol.

Can health kiosks be used to identify oral health care needs? A pilot study

Objective The aim of this study was to investigate the reliability of digital imaging for detecting restorative treatment need among individuals in their 20s by comparing the outcome of digital imaging with clinical caries findings at the patient level. Material and methods Five intraoral clinical daylight and digital fluorescence images were taken extraorally of 21 patients. A clinical examination was then performed by a trained and calibrated dentist. Additionally, the patients answered a multiple-choice questionnaire about their health habits. The images were analysed and caries findings were recorded. For statistical analysis, sensitivity and specificity were calculated. Results were shown as ROC curves and AUC values. All analyses were done using SPSS (version 24.0, Chicago, IL). Results Caries lesions were most often detected in molars and least often in canines. When using the clinical status as gold standard, digital imaging gave an AUC value of 0.617, whereas the outcome by questionnaire gave an AUC value of 0.719. When using the combined outcome of digital imaging and the questionnaire, the AUC value was 0.694 with clinical validation. Conclusions It can be concluded that health kiosks may help to reduce the number of patients waiting for dental treatment; more specifically, the questionnaire with individual feedback may provide a new instrument for providing instructions for homecare online. However, the camera system must be developed further, and dentists and dental hygienists require training to analyse the images.

Glomerular Classification Using Convolutional Neural Networks Based on Defined Annotation Criteria and Concordance Evaluation Among Clinicians

IntroductionDiagnosing renal pathologies is important for performing treatments. However, classifying every glomerulus is difficult for clinicians; thus, a support system, such as a computer, is required. This paper describes the automatic classification of glomerular images using a convolutional neural network (CNN).MethodTo generate appropriate labeled data, annotation criteria including 12 features (e.g., “fibrous crescent”) were defined. The concordance among 5 clinicians was evaluated for 100 images using the kappa (κ) coefficient for each feature. Using the annotation criteria, 1 clinician annotated 10,102 images. We trained the CNNs to classify the features with an average κ ≥0.4 and evaluated their performance using the receiver operating characteristic–area under the curve (ROC–AUC). An error analysis was conducted and the gradient-weighted class activation mapping (Grad-CAM) was also applied; it expresses the CNN’s focusing point with a heat map when the CNN classifies the glomerular image for a feature.ResultsThe average κ coefficient of the features ranged from 0.28 to 0.50. The ROC–AUC of the CNNs for test data varied from 0.65 to 0.98. Among the features, “capillary collapse” and “fibrous crescent” had high ROC–AUC values of 0.98 and 0.91, respectively. The error analysis and the Grad-CAM visually showed that the CNN could not distinguish between 2 different features that had similar visual structures or that occurred simultaneously.ConclusionThe differences in the texture or frequency of the co-occurrence between the different features affected the CNN performance; thus, to improve the classification accuracy, methods such as segmentation are required.

A machine learning-based method for prediction of macrocyclization patterns of polyketides and non-ribosomal peptides.

Even though genome mining tools have successfully identified large numbers of non-ribosomal peptide synthetase (NRPS) and polyketide synthase (PKS) biosynthetic gene clusters (BGCs) in bacterial genomes, currently no tool can predict the chemical structure of the secondary metabolites biosynthesized by these BGCs. Lack of algorithms for predicting complex macrocyclization patterns of linear PK/NRP biosynthetic intermediates has been the major bottleneck in deciphering the final bioactive chemical structures of PKs/NRPs by genome mining. Using a large dataset of known chemical structures of macrocyclized PKs/NRPs, we have developed a machine learning (ML) algorithm for distinguishing the correct macrocyclization pattern of PKs/NRPs from the library of all theoretically possible cyclization patterns. Benchmarking of this ML classifier on completely independent datasets has revealed ROC-AUC and PR-AUC values of 0.82 and 0.81, respectively. This cyclization prediction algorithm has been used to develop SBSPKSv3, a genome mining tool for completely automated prediction of macrocyclized structures of NRPs/PKs. SBSPKSv3 has been extensively benchmarked on a dataset of over 100 BGCs with known PKs/NRPs products. The macrocyclization prediction pipeline and all the datasets used in this study are freely available at http://www.nii.ac.in/sbspks3.html. Supplementary data are available at Bioinformatics online.

Comparative performance assessment of landslide susceptibility models with presence-only, presence-absence, and pseudo-absence data

The quality of the data for statistical methods plays an important role in landslide susceptibility mapping. How different data types influence the performance of landslide susceptibility maps is worth studying. The goal of this study was to explore the effects of different data types namely, presence-only (PO), presence-absence (PA), and pseudo-absence (PAs) data, on the predictive capability of landslide susceptibility mapping. This was completed by conducting a case study in the landslide-prone Honghe County in the Yunnan Province of China. A total of 428 landslide PO data points were selected. An equivalent number of non-landslide locations were generated as PA data by random sampling, and 10,000 sites were uniformly selected at random from each region as PAs data. Three landslide susceptibility models, namely the information value model (IVM), logistic regression (LR) model, and maximum entropy (MaxEnt) model, corresponding to the three data types were investigated. Additionally, the area under the receiver operating characteristic curves (ROC-AUC), seven statistical indices (i.e. accuracy, sensibility, false-positive rate, specificity, precision, Kappa, and F-measure), and a landslide density analysis were used to evaluate model performance regarding landslide susceptibility mapping. Our results indicated that the MaxEnt model using PAs data performed the best and had the highest fitness with the highest ROC-AUC values and statistical indices, followed by the IVM model with only landslide data (PO), and the LR model using PA data. Using PAs data avoided the inherent over-predictive shortcomings of PO data by limiting the predicted area of high-landslide susceptibility. Additionally, the random sampling design of landslide PA data increased the uncertainty of landslide susceptibility mapping and influenced the performance of the model. Therefore, our results suggested that the PAs data sampling provided a useful data type in the absence of high-quality data. Finally, we summarized the principles, advantages, and disadvantages of the three data types to assist with model optimization and the improvement of predicted performance and fitness.

Localized prediction of tissue outcome in acute ischemic stroke patients using diffusion- and perfusion-weighted MRI datasets.

BackgroundAn accurate prediction of tissue outcome in acute ischemic stroke patients is of high interest for treatment decision making. To date, various machine learning models have been proposed that combine multi-parametric imaging data for this purpose. However, most of these machine learning models were trained using voxel information extracted from the whole brain, without taking differences in susceptibility to ischemia into account that exist between brain regions. The aim of this study was to develop and evaluate a local tissue outcome prediction approach, which makes predictions using locally trained machine learning models and thus accounts for regional differences.Material and methodsMulti-parametric MRI data from 99 acute ischemic stroke patients were used for the development and evaluation of the local tissue outcome prediction approach. Diffusion (ADC) and perfusion parameter maps (CBF, CBV, MTT, Tmax) and corresponding follow-up lesion masks for each patient were registered to the MNI brain atlas. Logistic regression (LR) and random forest (RF) models were trained employing a local approach, which makes predictions using models individually trained for each specific voxel position using the corresponding local data. A global approach, which uses a single model trained using all voxels of the brain, was used for comparison. Tissue outcome predictions resulting from the global and local RF and LR models, as well as a combined (hybrid) approach were quantitatively evaluated and compared using the area under the receiver operating characteristic curve (ROC AUC), the Dice coefficient, and the sensitivity and specificity metrics.ResultsStatistical analysis revealed the highest ROC AUC and Dice values for the hybrid approach. With 0.872 (ROC AUC; LR) and 0.353 (Dice; RF), these values were significantly higher (p < 0.01) than the values of the two other approaches. In addition, the local approach achieved the highest sensitivity of 0.448 (LR). Overall, the hybrid approach was only outperformed in sensitivity (LR) by the local approach and in specificity by both other approaches. However, in these cases the effect sizes were comparatively small.ConclusionThe results of this study suggest that using locally trained machine learning models can lead to better lesion outcome prediction results compared to a single global machine learning model trained using all voxel information independent of the location in the brain.

Cardiac vagal tone as a novel screening tool to recognize asymptomatic cardiovascular autonomic neuropathy: Aspects of utility in type 1 diabetes

AimsTo test the performance of the cardiac vagal tone (CVT) derived from a 5-minute ECG recording compared with the standardized cardiovascular autonomic reflex tests (CARTs). MethodsCross-sectional study included 56 well-phenotyped adults with type 1 diabetes (19–71 years, 2–54 years disease-duration). Autonomic testing included: standardized CARTs obtained with the VAGUS™, CVT, and indices of heart rate variability (HRV) obtained at 24- and 120-hour, and electrochemical skin conductance assessed with SUDOSCAN®. ROC AUC and cut-off values were calculated for CVT to recognize CAN based on ≥ 2 (established CAN, n = 7) or 1 (borderline CAN, n = 9) abnormal CARTs and compared to HRV indices and electrochemical skin conductance. ResultsEstablished CAN: The cut-off CVT value of 3.2LVS showed 67% sensitivity and 87% specificity (p = 0.01). Indices of HRV at either 24-hour (AUC > 0.90) and 120-hour (AUC > 0.88) performed better than CVT. Borderline CAN: The cut-off CVT value of 5.2LVS indicated 88% sensitivity and 63% specificity (p = 0.07). CVT performed better than HRV indices (AUC < 0.72). Electrochemical skin conductance (AUC:0.63–0.72) had lower sensitivity and specificity compared with CVT. ConclusionsImplementation of CVT with a clinically applicable cut-off value may be considered a quicker and accessible screening tool which could ultimately decrease the number of unrecognized CAN and initiate earlier prevention initiatives.

Liver Steatosis Categorization on Contrast-Enhanced CT Using a Fully Automated Deep Learning Volumetric Segmentation Tool: Evaluation in 1204 Healthy Adults Using Unenhanced CT as a Reference Standard.

BACKGROUND. Hepatic attenuation at unenhanced CT is linearly correlated with the MRI proton density fat fraction (PDFF). Liver fat quantification at contrast-enhanced CT is more challenging. OBJECTIVE. The purpose of this article is to evaluate liver steatosis categorization on contrast-enhanced CT using a fully automated deep learning volumetric hepatosplenic segmentation algorithm and unenhanced CT as the reference standard. METHODS. A fully automated volumetric hepatosplenic segmentation algorithm using 3D convolutional neural networks was applied to unenhanced and contrast-enhanced series from a sample of 1204 healthy adults (mean age, 45.2 years; 726 women, 478 men) undergoing CT evaluation for renal donation. The mean volumetric attenuation was computed from all designated liver and spleen voxels. PDFF was estimated from unenhanced CT attenuation and served as the reference standard. Contrast-enhanced attenuations were evaluated for detecting PDFF thresholds of 5% (mild steatosis, 10% and 15% (moderate steatosis); PDFF less than 5% was considered normal. RESULTS. Using unenhanced CT as reference, estimated PDFF was ≥ 5% (mild steatosis), ≥ 10%, and ≥ 15% (moderate steatosis) in 50.1% (n = 603), 12.5% (n = 151) and 4.8% (n = 58) of patients, respectively. ROC AUC values for predicting PDFF thresholds of 5%, 10%, and 15% using contrast-enhanced liver attenuation were 0.669, 0.854, and 0.962, respectively, and using contrast-enhanced liver-spleen attenuation difference were 0.662, 0.866, and 0.986, respectively. A total of 96.8% (90/93) of patients with contrast-enhanced liver attenuation less than 90 HU had steatosis (PDFF ≥ 5%); this threshold of less than 90 HU achieved sensitivity of 75.9% and specificity of 95.7% for moderate steatosis (PDFF ≥ 15%). Liver attenuation less than 100 HU achieved sensitivity of 34.0% and specificity of 94.2% for any steatosis (PDFF ≥ 5%). A total of 93.8% (30/32) of patients with contrast-enhanced liver-spleen attenuation difference 10 HU or less had moderate steatosis (PDFF ≥ 15%); a liver-spleen difference less than 5 HU achieved sensitivity of 91.4% and specificity of 95.0% for moderate steatosis. Liver-spleen difference less than 10 HU achieved sensitivity of 29.5% and specificity of 95.5% for any steatosis (PDFF ≥ 5%). CONCLUSION. Contrast-enhanced volumetric hepatosplenic attenuation derived using a fully automated deep learning CT tool may allow objective categoric assessment of hepatic steatosis. Accuracy was better for moderate than mild steatosis. Further confirmation using different scanning protocols and vendors is warranted. CLINICAL IMPACT. If these results are confirmed in independent patient samples, this automated approach could prove useful for both individualized and population-based steatosis assessment.

IDSSIM: an lncRNA functional similarity calculation model based on an improved disease semantic similarity method

BackgroundIt has been widely accepted that long non-coding RNAs (lncRNAs) play important roles in the development and progression of human diseases. Many association prediction models have been proposed for predicting lncRNA functions and identifying potential lncRNA-disease associations. Nevertheless, among them, little effort has been attempted to measure lncRNA functional similarity, which is an essential part of association prediction models.ResultsIn this study, we presented an lncRNA functional similarity calculation model, IDSSIM for short, based on an improved disease semantic similarity method, highlight of which is the introduction of information content contribution factor into the semantic value calculation to take into account both the hierarchical structures of disease directed acyclic graphs and the disease specificities. IDSSIM and three state-of-the-art models, i.e., LNCSIM1, LNCSIM2, and ILNCSIM, were evaluated by applying their disease semantic similarity matrices and the lncRNA functional similarity matrices, as well as corresponding matrices of human lncRNA-disease associations coming from either lncRNADisease database or MNDR database, into an association prediction method WKNKN for lncRNA-disease association prediction. In addition, case studies of breast cancer and adenocarcinoma were also performed to validate the effectiveness of IDSSIM.ConclusionsResults demonstrated that in terms of ROC curves and AUC values, IDSSIM is superior to compared models, and can improve accuracy of disease semantic similarity effectively, leading to increase the association prediction ability of the IDSSIM-WKNKN model; in terms of case studies, most of potential disease-associated lncRNAs predicted by IDSSIM can be confirmed by databases and literatures, implying that IDSSIM can serve as a promising tool for predicting lncRNA functions, identifying potential lncRNA-disease associations, and pre-screening candidate lncRNAs to perform biological experiments. The IDSSIM code, all experimental data and prediction results are available online at https://github.com/CDMB-lab/IDSSIM.

Exploring technical issues in personalized medicine: NSCLC survival prediction by quantitative image analysis-usefulness of density correction of volumetric CT data.

The aim of this study was to apply density correction method to the quantitative image analysis of non-small cell lung cancer (NSCLC) computed tomography (CT) images, determining its influence on overall survival (OS) prediction of surgically treated patients. Clinicopathological (CP) data and preoperative CT scans, pre- and post-contrast medium (CM) administration, of 57 surgically treated NSCLC patients, were retrospectively collected. After CT volumetric density measurement of primary gross tumour volume (GTV), aorta and tracheal air, density correction was conducted on GTV (reference values: aortic blood and tracheal air). For each resulting data set (combining CM administration and normalization), first-order statistical and textural features were extracted. CP and imaging data were correlated with patients 1-, 3- and 5-year OS, alone and combined (uni-/multivariate logistic regression and Akaike information criterion). Predictive performance was evaluated using the ROC curves and AUC values and compared among non-normalized/normalized data sets (DeLong test). The best predictive values were obtained when combining CP and imaging parameters (AUC values: 1year 0.72; 3years 0.82; 5years 0.78). After normalization resulted an improvement in predicting 1-year OS for some of the grey level size zonebased features (large zone low grey level emphasis) and for the combined CP-imaging model, a worse performance for grey level co-occurrence matrix (cluster prominence and shade) and first-order statistical (range) parameters for 1- and 5-year OS, respectively. The negative performance of cluster prominence in predicting 1-year OS was the only statistically significant result (p value 0.05). Density corrections of volumetric CT data showed an opposite influence on the performance of imaging quantitative features in predicting OS of surgically treated NSCLC patients, even if no statistically significant for almost all predictors.

Identification of Switch-Like Features in Proteins using Sequence-Based Descriptors Modelled from Regions Identified in X-Ray Crystal Structures

Identifying possible associated switches from amino acid sequence is likely of much utility. We predict possible switch-like regions from protein sequences using a logistic regression approach with forty-eight acetyltransferases as a learning set functionally related to the five sirtuins of our validation set. Validated residue binary assignments for secondary structure change were determined (DSSP) from 3D X-ray structures for sets of virtually identical chains crystallized under different conditions. Some 1796 residue positions were identified as structurally variable in the 13,797-residue learning set as opposed to 436 such residues in the 2754-residue test set. Our sequence descriptors include amino acid type, 6 and 20-term sequence entropy, Lobanov-Galzitskaya residue disorder propensity, Vkabat (variability with respect to predictions from sequence of helix, sheet and other), and all possible combinations. We find the optimal AUC values approaching 70% for two separate models involving either residue disorder or Vkabat. An additional descriptor involving charge did not significantly improve prediction for the test set. Calculations did suggest a significant fraction of possible switch-like regions as being adjacent to the allosteric region for Human Sirt-1. Analysis on the larger learning set indicated the best models included both the residue disorder and Vkabat descriptors, as well as 20-term sequence entropy and charge, with residue disorder being most the most significant descriptor. Note some 2984 aligned-residue positions, determined from sets of 3D-structures for homologues under different conditions, included at least one aligned-residue that did not have an assigned secondary structure. Predictions of switch-like residues descriptors for such unassigned residues showed improved ROC AUC values exceeding 90% when applying the residue disorder descriptor. Further characterization of these unassigned residues is suggested to better classify residues as specifically disordered.

The CAHP (cardiac arrest hospital prognosis) score: A tool for risk stratification after out-of-hospital cardiac arrest in elderly patients

BackgroundOlder age is associated with worse outcome after out-of-hospital cardiac arrest (OHCA). Therefore, we tested the performance of CAHP score, to predict neurological outcome in elderly OHCA patients and to select patients most likely to benefit from coronary angiogram (CAG). Materials and methodsThe present study was a retrospective multicentre observational study at 3 non-university hospitals and 1 university hospital. CAHP score was calculated, and its performance to predict outcomes was evaluated. Factors associated with the use of CAG were analysed and the rate of CAG across each CAHP score risk group reported. ResultsOne hundred seventy-six patients fulfilled inclusion criteria (median age of 81, [79–84]), among which a cardiac cause was presumed for 99 patients. The hospital unfavourable outcome was 91%. The ROC-AUC values for hospital neurological outcome prediction of CAHP score was 0.81 [0.68–0.94], showing good discrimination performance. ST-segment elevation in ECG and initial shockable rhythm were independent factors for performing early CAG, whereas age and distance from the percutaneous coronary intervention centre were independently associated with the absence of early CAG. The percentages of patients receiving early CAG in the low, medium and high CAHP score risk groups were 64%, 33% and 34%, respectively, and differed significantly between low CAHP score risk group and other groups (p = 0.02). ConclusionsThe CAHP score exhibited a good discrimination performance to predict neurological outcome in elderly OHCA patients. This score could represent a helpful tool for treatment allocation. A simple prognostication score could permit avoiding unnecessary procedures in patients with minimal chances of survival.

Burns objective scar scale (BOSS): Validation of an objective measurement devices based burn scar scale panel

AimsHypertrophic scars in burn survivors are a major cause of morbidity but the development of evidence based treatments is hampered by the lack of objective measurements of these scars. The objective of our study is to investigate the most accurate parameters for objective scar assessment and to create a combination score to facilitate the use of a panel of objective scar measurement tools. MethodsThree independent assessors evaluated fifty five scar sites on fifty five burn patients with both the subjective modified Vancouver Scar Scale (mVSS) and a panel of objective measurement tools including the DSM II Colormeter, Cutometer, Dermascan high frequency ultrasound. The sensitivity and specificity of the objective scar parameters in predicting a mVSS score of 6 or more using the Receiving Operator Characteristic Area under the curve (ROC AUC) was then calculated and the most accurate parameters were combined to create an objective global scar score. ResultsThe ROC AUC values were found to be highest for the Dermascan scar thickness (0.897), dermal intensity and intensity ratio (0.914 and 0.919), Cutometer R0 value (0.942), and R0 ratio (0.944). For colour measurements, ratios of scar to normal skin performed better than the single parameters for both erythema and pigmentation measurements: DSM II Erythema ratio vs Erythema (0.885 vs 0.818), DSM II a* ratio vs a* (0.848 vs 0.741); DSM II Melanin ratio vs Melanin (0.854 vs 0.761), DSM II L* ratio vs L* (0.862 vs 0.767). Analysis of the ROC AUC with chi-square test values showed that the highest AUC (0.786) was obtained with the combination of the Cutometer R0, Dermascan scar thickness, intensity and their respective scar to normal skin ratios. A total score of 5 and above (out of 6 parameters) had the highest combined sensitivity (69.0%) and specificity (83.3%). ConclusionThe objective parameters for the DSM II Colormeter, Cutometer and Dermascan high frequency ultrasound were all found to have moderate to strong ROC AUC values and combination of the Cutometer R0 and Dermascan scar thickness and intensity values can be used to create an objective global scar scale that can accurately differentiate patients with hypertrophic burn scarring from non-hypertrophic scars or normal skin.

Differentiating Pancreatic Ductal Adenocarcinoma (PDAC) from individuals with symptoms suggestive of PDAC, including type II diabetes, with ROC AUC values above 0.95

Universal health care (UHC) should ensure equal access to and use of surgery, but few studies have explored variation in UHC systems. The objective was to describe practice of distal pancreatectomy in Norway covered exclusively by an UHC.Data on all patients undergoing distal pancreatectomy from the Norwegian Patient Register over a 5-year period. Age- and gender-adjusted population-based resection rates (adj. per million/yr) for distal pancreatectomy were analysed across 4 regions and outcomes related to splenic salvage rate, hospital stay, reoperation, readmissions and 90-day mortality risk between regions. Risk is reported as odds ratio (OR) with 95% confidence interval (c.i.).Regional difference exist in terms of absolute numbers, with the majority of procedures done in one region (n = 331; 59.7%). Regional variation persisted for age- and gender-adjusted population-rates, with highest rate at 23.8/million/yr and lowest rate at 13.5/mill/yr (for a 176% relative difference; or an absolute difference of +10.3 resections/million/yr). Overall, a lapDP instead of an open DP was 3.5 times more likely in SouthEast compared to all other regions combined (lapDP rate: 83% vrs 24%, respectively; OR 15.4, 95% c.i. 10.1–23.5; P < 0.001). The splenic salvage rate was lower in SouthEast (19.9%) compared to all other regions (average 26.5%; highest in Central-region at 37.0%; P = 0.010 for trend). Controlled for other factors in multivariate regression, ‘region’ of surgery remained significantly associated with laparoscopic access.Despite a universal health care system, considerable variation exists in resection rates, use of laparoscopy and splenic salvage rates across regions.