Prediction of MHC class II-binding peptides using an evolutionary algorithm and artificial neural network

Motivation: Prediction methods for identifying binding peptides could minimize the number of peptides required to be synthesized and assayed, and thereby facilitate the identification of potential T-cell epitopes. We developed a bioinformatic method for the prediction of peptide binding to MHC class II molecules. Results: Experimental binding data and expert knowledge of anchor positions and binding motifs were combined with an evolutionary algorithm (EA) and an artificial neural network (ANN): binding data extraction --> peptide alignment --> ANN training and classification. This method, termed PERUN, was implemented for the prediction of peptides that bind to HLA-DR4(B1*0401). The respective positive predictive values of PERUN predictions of high-, moderate-, low- and zero-affinity binder-a were assessed as 0.8, 0.7, 0.5 and 0.8 by cross-validation, and 1.0, 0.8, 0.3 and 0.7 by experimental binding. This illustrates the synergy between experimentation and computer modeling, and its application to the identification of potential immunotheraaeutic peptides.

Computational immunology: The coming of age

The explosive growth in biotechnology combined with major advancesin information technology has the potential to radically transformimmunology in the postgenomics era. Not only do we now have readyaccess to vast quantities of existing data, but new data with relevanceto immunology are being accumulated at an exponential rate. Resourcesfor computational immunology include biological databases and methodsfor data extraction, comparison, analysis and interpretation. Publiclyaccessible biological databases of relevance to immunologists numberin the hundreds and are growing daily. The ability to efficientlyextract and analyse information from these databases is vital forefficient immunology research. Most importantly, a new generationof computational immunology tools enables modelling of peptide transportby the transporter associated with antigen processing (TAP), modellingof antibody binding sites, identification of allergenic motifs andmodelling of T-cell receptor serial triggering.

Whispered voice test for screening for hearing impairment in adults and children: systematic review

Objective To determine the accuracy of the whispered voice test in detecting hearing impairment in adults and children. Design Systematic review of studies of test accuracy. Data sources Medline, Embase, Science Citation Index, unpublished theses, manual searching of bibliographies of known primary and review articles, and contact with authors. Study selection Two reviewers independently selected and extracted data on study characteristics, quality, and accuracy of studies. Studies were included if they had cross sectional designs, at least one of the index tests was the whispered voice test, and the reference test (audiometry) was performed on at least 80% of the participants. Data extraction Data were used to form 2x2 contingency tables with hearing impairment by audiometry as the reference standard. Data synthesis The eight studies that were found used six different techniques. The sensitivity in the four adult studies was 90% or 100% and the specificity was 70% to 87%. The sensitivity in the four childhood studies ranged from 80% to 96% and specificity ranged from 90% to 98%. Conclusion The whispered voice test is a simple and accurate test for detecting hearing impairment. There is some concern regarding the lower sensitivity in children and the overall reproducibility of the test, particularly in primary care settings. Further studies should be conducted in primary care settings to explore the influence of components of the testing procedure to optimise test sensitivity and to promote standardisation of the testing procedure.

How well does B-type natriuretic peptide predict death and cardiac events in patients with heart failure: systematic review

Objective To assess how well B-type natriuretic peptide (BNP) predicts prognosis in patients with heart failure. Design Systematic review of studies assessing BNP for prognosis m patients with heart failure or asymptomatic patients. Data sources Electronic searches of Medline and Embase from January 1994 to March 2004 and reference lists of included studies. Study selection and data extraction We included all studies that estimated the relation between BNP measurement and the risk of death, cardiac death, sudden death, or cardiovascular event in patients with heart failure or asymptomatic patients, including initial values and changes in values in response to treatment. Multivariable models that included both BNP and left ventricular ejection fraction as predictors were used to compare the prognostic value of each variable. Two reviewers independently selected studies and extracted data. Data synthesis 19 studies used BNP to estimate the relative risk of death or cardiovascular events in heart failure patients and five studies in asymptomatic patients. In heart failure patients, each 100 pg/ml increase was associated with a 35% increase in the relative risk of death. BNP was used in 35 multivariable models of prognosis. In nine of the models, it was the only variable to reach significance-that is, other variables contained no prognostic information beyond that of BNP. Even allowing for the scale of the variables, it seems to be a strong indicator of risk. Conclusion Although systematic reviews of prognostic studies have inherent difficulties, including die possibility of publication bias, the results of the studies in this review show that BNP is a strong prognostic indicator for both asymptomatic patients mid for patients with heart failure at all stages of disease.

Lipid-lowering therapy following major cardiac events: progress and deficits

Objective: To assess hospital prescribing of lipid-lowering agents in a tertiary hospital, and examine continuation of, or changes to, such therapy in the 6-18 months following discharge. Design: Retrospective data extraction from the hospital records of patients admitted from October 1998 to April 1999. These patients and their general practitioners were then contacted to obtain information about ongoing management after discharge. Setting: Tertiary public hospital and community. Participants: 352 patients admitted to hospital with acute myocardial infarction or unstable angina, and their GPs. Main outcome measures: Percentage of eligible patients discharged on lipid-lowering therapy and percentage of patients continuing or starting such therapy 6-18 months after discharge. Results: 10% of inpatients with acute coronary syndromes did not have lipid-level estimations performed or arranged during admission. Documentation of lipid levels in discharge summaries was poor. Eighteen per cent of patients with a total serum cholesterol level greater than 5.5 mmol/L did not receive a discharge prescription for a cholesterol-lowering agent. Compliance with treatment on follow-up was 88% in the group discharged on treatment. However, at follow-up, 70% of patients discharged without therapy had not been commenced on lipid-lowering treatment by their GPs. Conclusions: Prescribing of lipid-lowering therapy for secondary prevention following acute coronary syndromes remains suboptimal. Commencing treatment in hospital is likely to result in continuing therapy in the community. Better communication of lipid-level results, treatment and treatment aims between hospitals and GPs might encourage optimal treatment practices.

Cognitive style factors affecting database query performance

As end-user computing becomes more pervasive, an organization's success increasingly depends on the ability of end-users, usually in managerial positions, to extract appropriate data from both internal and external sources. Many of these data sources include or are derived from the organization's accounting information systems. Managerial end-users with different personal characteristics and approaches are likely to compose queries of differing levels of accuracy when searching the data contained within these accounting information systems. This research investigates how cognitive style elements of personality influence managerial end-user performance in database querying tasks. A laboratory experiment was conducted in which participants generated queries to retrieve information from an accounting information system to satisfy typical information requirements. The experiment investigated the influence of personality on the accuracy of queries of varying degrees of complexity. Relying on the Myers–Briggs personality instrument, results show that perceiving individuals (as opposed to judging individuals) who rely on intuition (as opposed to sensing) composed queries more accurately. As expected, query complexity and academic performance also explain the success of data extraction tasks.

Risk of serious NSAID-related gastrointestinal events during long-term exposure: a systematic review

Objective: Exposure to non-steroidal anti-inflammatory drugs (NSAIDs) is associated wit increased risk of serious gastrointestinal (GI) events compared with non-exposure. We investigated whether that risk is sustained over time. Data sources: Cochrane Controlled Trials Register (to 2002); MEDLINE, EMBASE, Derwent Drug File and Current Contents (1999-2002); manual searching of reviews (1999-2002). Study selection: From 479 search results reviewed and 221 articles retrieved, seven studies of patients exposed to prescription non-selective NSAIDs for more than 6 months and reporting time-dependent serious GI event rates were selected for quantitative data synthesis. These were stratified into two groups by study design. Data extraction: Incidence of GI events and number of patients at specific time points were extracted. Data synthesis: Meta-regression analyses were performed. Change in risk was evaluated by testing whether the slope of the regression line declined over time. Four randomised controlled trials (RCTs) provided evaluable data from five NSAID arms (aspirin, naproxen, two ibuprofen arms, and diclofenac). When the RCT data were combined, a small significant decline in annualised risk was seen: -0.005% (95% Cl, -0.008% to -0.001%) per month. Sensitivity analyses were conducted because there was disparity within the RCT data. The pooled estimate from three cohort studies showed no significant decline in annualised risk over periods up to 2 years: -0.003% (95% Cl, -0.008% to 0.003%) per month. Conclusions: Small decreases in risk over time were observed; these were of negligible clinical importance. For patients who need long-term (> 6 months) treatment, precautionary measures should be considered to reduce the net probability of serious GI events over the anticipated treatment duration. The effect of intermittent versus regular daily therapy on long-term risk needs further investigation.

Radiofrequency ablation of liver tumors - A systematic review

Objectives: To systematically review radiofrequency ablation (RFA) for treating liver tumors. Data Sources: Databases were searched in July 2003. Study Selection: Studies comparing RFA with other therapies for hepatocellular carcinoma (HCC) and colorectal liver metastases (CLM) plus selected case series for CLM. Data Extraction: One researcher used standardized data extraction tables developed before the study, and these were checked by a second researcher. Data Synthesis: For HCC, 1.3 comparative studies were included, 4 of which were randomized, controlled trials. For CLM, 13 studies were included, 2 of which were nonrandomized comparative studies and 11 that were case series. There did not seem to be any distinct differences in the complication rates between RFA and any of the other procedures for treatment of HCC. The local recurrence rate at 2 years showed a statistically significant benefit for RFA over percutaneous ethanol injection for treatment of HCC (6% vs 26%, 1 randomized, controlled trial). Local recurrence was reported to be more common after RFA than after laser-induced thermotherapy, and a higher recurrence rate and a shorter time to recurrence were dassociated with RFA compared with surgical resection (1 nonrandomized study each). For CLM, the postoperative complication rate ranged from 0% to 33% (3 case series). Survival after diagnosis was shorter in the CLM group treated with RFA than in the surgical resection group (1 nonrandomized study). The CLM local recurrence rate after RFA ranged from 4% to 55% (6 case series). Conclusions: Radiofrequency ablation may be more effective than other treatments in terms of less recurrence of HCC and may be as sale, although the evidence is scant. There was not enough evidence to determine the safety or efficacy of RFA for treatment of CLM.

Selection bias in gene extraction on the basis of microarray gene-expression data

In the context of cancer diagnosis and treatment, we consider the problem of constructing an accurate prediction rule on the basis of a relatively small number of tumor tissue samples of known type containing the expression data on very many (possibly thousands) genes. Recently, results have been presented in the literature suggesting that it is possible to construct a prediction rule from only a few genes such that it has a negligible prediction error rate. However, in these results the test error or the leave-one-out cross-validated error is calculated without allowance for the selection bias. There is no allowance because the rule is either tested on tissue samples that were used in the first instance to select the genes being used in the rule or because the cross-validation of the rule is not external to the selection process; that is, gene selection is not performed in training the rule at each stage of the cross-validation process. We describe how in practice the selection bias can be assessed and corrected for by either performing a cross-validation or applying the bootstrap external to the selection process. We recommend using 10-fold rather than leave-one-out cross-validation, and concerning the bootstrap, we suggest using the so-called. 632+ bootstrap error estimate designed to handle overfitted prediction rules. Using two published data sets, we demonstrate that when correction is made for the selection bias, the cross-validated error is no longer zero for a subset of only a few genes.

Relative dispersions of intra-albumin transit times across rat and elasmobranch perfused livers, and implications for intra- and inter-species scaling of hepatic clearance using microsomal data

It is recognized that vascular dispersion in the liver is a determinant of high first-pass extraction of solutes by that organ. Such dispersion is also required for translation of in-vitro microsomal activity into in-vivo predictions of hepatic extraction for any solute. We therefore investigated the relative dispersion of albumin transit times (CV2) in the livers of adult and weanling rats and in elasmobranch livers. The mean and normalized variance of the hepatic transit time distribution of albumin was estimated using parametric non-linear regression (with a correction for catheter influence) after an impulse (bolus) input of labelled albumin into a single-pass liver perfusion. The mean +/- s.e. of CV2 for albumin determined in each of the liver groups were 0.85 +/- 0.20 (n = 12), 1.48 +/- 0.33 (n = 7) and 0.90 +/- 0.18 (n = 4) for the livers of adult and weanling rats and elasmobranch livers, respectively. These CV2 are comparable with that reported previously for the dog and suggest that the CV2 Of the liver is of a similar order of magnitude irrespective of the age and morphological development of the species. It might, therefore, be justified, in the absence of other information, to predict the hepatic clearances and availabilities of highly extracted solutes by scaling within and between species livers using hepatic elimination models such as the dispersion model with a CV2 of approximately unity.

A reproducible method for automated extraction of brain volumes from 3D human head MR images

An automated method for extracting brain volumes from three commonly acquired three-dimensional (3D) MR images (proton density, T1 weighted, and T2-weighted) of the human head is described. The procedure is divided into four levels: preprocessing, segmentation, scalp removal, and postprocessing. A user-provided reference point is the sole operator-dependent input required, The method's parameters were first optimized and then fixed and applied to 30 repeat data sets from 15 normal older adult subjects to investigate its reproducibility. Percent differences between total brain volumes (TBVs) for the subjects' repeated data sets ranged from .5% to 2.2%. We conclude that the method is both robust and reproducible and has the potential for wide application.

Identification of "pathologs" (disease-related genes) from the RIKEN mouse cDNA dataset using human curation plus FACTS, a new biological information extraction system

Background: A major goal in the post-genomic era is to identify and characterise disease susceptibility genes and to apply this knowledge to disease prevention and treatment. Rodents and humans have remarkably similar genomes and share closely related biochemical, physiological and pathological pathways. In this work we utilised the latest information on the mouse transcriptome as revealed by the RIKEN FANTOM2 project to identify novel human disease-related candidate genes. We define a new term patholog to mean a homolog of a human disease-related gene encoding a product ( transcript, anti-sense or protein) potentially relevant to disease. Rather than just focus on Mendelian inheritance, we applied the analysis to all potential pathologs regardless of their inheritance pattern. Results: Bioinformatic analysis and human curation of 60,770 RIKEN full-length mouse cDNA clones produced 2,578 sequences that showed similarity ( 70 - 85% identity) to known human-disease genes. Using a newly developed biological information extraction and annotation tool ( FACTS) in parallel with human expert analysis of 17,051 MEDLINE scientific abstracts we identified 182 novel potential pathologs. Of these, 36 were identified by computational tools only, 49 by human expert analysis only and 97 by both methods. These pathologs were related to neoplastic ( 53%), hereditary ( 24%), immunological ( 5%), cardio-vascular (4%), or other (14%), disorders. Conclusions: Large scale genome projects continue to produce a vast amount of data with potential application to the study of human disease. For this potential to be realised we need intelligent strategies for data categorisation and the ability to link sequence data with relevant literature. This paper demonstrates the power of combining human expert annotation with FACTS, a newly developed bioinformatics tool, to identify novel pathologs from within large-scale mouse transcript datasets.

Extraction of electrical characteristics from pixels of multifrequency EIT images

Computer modelling has shown that electrical characteristics of individual pixels may be extracted from within multiple-frequency electrical impedance tomography (MFEIT) images formed using a reference data set obtained from a purely resistive, homogeneous medium. In some applications it is desirable to extract the electrical characteristics of individual pixels from images where a purely resistive, homogeneous reference data set is not available. One such application of the technique of MFEIT is to allow the acquisition of in vivo images using reference data sets obtained from a non-homogeneous medium with a reactive component. However, the reactive component of the reference data set introduces difficulties with the extraction of the true electrical characteristics from the image pixels. This study was a preliminary investigation of a technique to extract electrical parameters from multifrequency images when the reference data set has a reactive component. Unlike the situation in which a homogenous, resistive data set is available, it is not possible to obtain the impedance and phase information directly from the image pixel values of the MFEIT images data set, as the phase of the reactive reference is not known. The method reported here to extract the electrical characteristics (the Cole-Cole plot) initially assumes that this phase angle is zero. With this assumption, an impedance spectrum can be directly extracted from the image set. To obtain the true Cole-Cole plot a correction must be applied to account for the inherent rotation of the extracted impedance spectrum about the origin, which is a result of the assumption. This work shows that the angle of rotation associated with the reactive component of the reference data set may be determined using a priori knowledge of the distribution of frequencies of the Cole-Cole plot. Using this angle of rotation, the true Cole-Cole plot can be obtained from the impedance spectrum extracted from the MFEIT image data set. The method was investigated using simulated data, both with and without noise, and also for image data obtained in vitro. The in vitro studies involved 32 logarithmically spaced frequencies from 4 kHz up to 1 MHz and demonstrated that differences between the true characteristics and those of the impedance spectrum were reduced significantly after application of the correction technique. The differences between the extracted parameters and the true values prior to correction were in the range from 16% to 70%. Following application of the correction technique the differences were reduced to less than 5%. The parameters obtained from the Cole-Cole plot may be useful as a characterization of the nature and health of the imaged tissues.