Statistical methods for the practical clinical application of morphometric measurements

Four statistical methods are presented to determine the practical clinical value of measurements made from malignant tumors and to translate these measurements into a prediction of survival for each patient: the Cox statistical model, which must be derived from a data base of cases with known outcome; the null-rank test, a modified rank-sum test that provides an overall measure of the effectiveness of the Cox model; the predicted survival curve, an estimate of survival derived for each new patient from measurements of the primary tumor; and the standard error of measurement, an empirical method for estimating the variability introduced into predicted survival by errors in measurement. The value of these statistical methods was demonstrated by application to 200 cases of human intraocular melanoma, with the two predictive morphometric measurements used being the standard deviation of nucleolar area (SDNA) and the largest tumor dimension (LTD) derived from a single histologic slide of each tumor. Sufficient references and mathematical details are provided to allow anyone with moderate skills as a computer programmer to construct or obtain all of the relevant algorithms.     

Accuracy and repeatability of moose (<Emphasis Type="Italic">Alces alces</Emphasis>) age as estimated from dental cement layers

Optimal research and management of species with age structure often depends on estimates of age-specific population parameters, which in turn depends on reliable methods for age determination. By counting annuli in the cementum of incisor root tips from 51 known-age moose (Alces alces) between 1 and 12 years old, we examined the variation in accuracy and repeatability of age estimates provided by three research technicians with different experiences of aging moose. The most experienced technician estimated the moose age correctly in up to 90% of the cases, while the technician with no prior experience estimated age correctly in up to 73% of the cases. The medium-experienced technician achieved a lower accuracy (up to 53%), indicating that experience alone is not sufficient if the basic training or lack of routine checks against other colleagues or a known-age material are not undertaken. The percentage moose aged within ±1 year from correct age was significantly higher in all technicians (94–98%). After the generally high accuracy, we also found high repeatability (0.980–0.994) within technicians. We conclude that this method of age-determination provides reliable estimates that can be used by management and research to gain information about age-specific patterns in moose populations. However, to obtain estimates of high accuracy the technicians should be well trained, have gained the necessary experience, and most preferably, have access to a sample of teeth from known-age moose.     

Assessing technician effects when extracting quantities from microscope images

Consider an experiment where the response is based on an image; e.g., an image captured to a computer file by a digital camera mounted on a microscope. Suppose relevant quantitative measures are extracted from the images so that results can be analyzed by conventional statistical methods. The steps involved in extracting the measures may require that the technicians, who are processing the images, perform some subjective manipulations. In this case, it is important to determine the bias and variability, if any, attributable to the technicians' decisions. This paper describes the experimental design and statistical analyses that are useful for those determinations. The design and analysis are illustrated by application to two biofilm research projects that involved quantitative image analysis. In one investigation, the technician was required to choose a threshold level, then the image analysis program automatically extracted relevant measures from the resulting black and white image. In the other investigation, the technician was required to choose fiducial points in each of two images collected on different microscopes; then the image analysis program registered the images by stretching, rotating, and overlaying them, so that their quantitative features could be correlated. These investigations elucidated the effects of the technicians' decisions, thereby helping us to assess properly the statistical uncertainties in the conclusions for the primary experiments.     

Quality assurance of canine vaginal cytology: A preliminary study

Regulatory controls of quality assurance in veterinary laboratories are less common than in human reproduction laboratories and the intra- and inter-technician variation in the assessment of canine vaginal cytology has not been reported. This study was designed to determine whether variation in classification of vaginal epithelial cells and interpretation of vaginal cytology smears existed within and between technicians in a canine reproductive laboratory.Sixteen vaginal cytology smears representing different known stages of the oestrous cycle were examined twice by one experienced technician and three inexperienced technicians in a blinded random order study design. Seven assessments were made; counting and classifying one hundred vaginal epithelial cells into four morphological classifications and assessment of three cellular categories. Technicians also interpreted their results and reported the stage of the cycle they thought each slide represented. In addition, selected samples were sent to four external commercial laboratories for interpretation.For the experienced technician, intra-technician variation was low for the morphological classifications and cellular assessments (r = 0.69-0.95). There was more intra-technician variation between results from Examination One and Examination Two for the inexperienced technicians (r = 0.53-0.92 where correlations were found). When inexperienced technicians' results were compared to results from the experienced technician, the inter-technician variation was low; results were correlated for 17 of the 21 observations (four morphological classifications and three cellular assessments across the three technicians) (r = 0.38-0.87). When technician interpretations of stage of the oestrous cycle were compared to the known stage of the cycle for each smear, the experienced technician correctly interpreted 19 of the 32 smears, whilst the three inexperienced technicians correctly interpreted 14, 16, and 18 of the 32 smears. The interpretation of vaginal smears by external laboratories was varied and sometimes inconclusive; 50% of laboratories incorrectly identified metoestrus smears as proestrus and 25% of the laboratories incorrectly identified an oestrus smear as proestrus.The results of this study are highly important for clinicians undertaking canine reproductive assessments since they demonstrate the potential for variability of results. While the greatest precision was found when vaginal smears were examined by an experienced technician (who, on a daily basis, examines many smears), more variability in both the reporting of different cell types and interpretation of the smears was observed by inexperienced technicians and when samples were sent to external commercial laboratories. These findings suggest that suitable quality control programmes should be implemented for laboratories that are undertaking routine assessments of canine reproductive function.     

Comparative study of automatic blood-gas analysers and their use in analysing arterial and capillary samples.

Three automatic blood-gas analysers were compared for ease of use; calibration; reproducibility and accuracy of results; maintenance; fault-finding; and use of expert technician time. Results obtained from arterial and capillary blood were compared with duplicate values obtained with a semi-automatic analyser controlled and calibrated with tonometered blood. No analyser was fully automatic, and all three needed maintenance by expert technicians. Difficulties were encountered when inexperienced operators used the machines. One automatic blood-gas analyser gave aberrant values for oxygen pressure (PO2) due to electrode dysfunction that was not indicated by the fault-finding system. A second analyser gave significantly lower values for blood pH than the standard machine. A comparison of pH, carbon dioxide pressure (PCO2), and PO2 measured in 40 simultaneous paired samples of arterial and arterialised capillary blood showed no significant difference for pH or PCO2, but the PO2 values were significantly lower in the capillary samples over the range studied. We conclude that all machines perform satisfactorily in terms of blood-gas analysis, but none may be regarded as fully automatic. Some degree of technical supervision is essential, as is proper training for all potential users.     

Reliability of Corneal Dynamic Scheimpflug Analyser Measurements in Virgin and Post-PRK Eyes

Purpose

To determine the measurement reliability of CorVis ST, a dynamic Scheimpflug analyser, in virgin and post-photorefractive keratectomy (PRK) eyes and compare the results between these two groups.

Methods

Forty virgin eyes and 42 post-PRK eyes underwent CorVis ST measurements performed by two technicians. Repeatability was evaluated by comparing three consecutive measurements by technician A. Reproducibility was determined by comparing the first measurement by technician A with one performed by technician B. Intraobserver and interobserver intraclass correlation coefficients (ICCs) were calculated. Univariate analysis of covariance (ANCOVA) was used to compare measured parameters between virgin and post-PRK eyes.

Results

The intraocular pressure (IOP), central corneal thickness (CCT) and 1_st applanation time demonstrated good intraobserver repeatability and interobserver reproducibility (ICC≧0.90) in virgin and post-PRK eyes. The deformation amplitude showed a good or close to good repeatability and reproducibility in both groups (ICC≧0.88). The CCT correlated positively with 1_st applanation time (r = 0.437 and 0.483, respectively, p<0.05) and negatively with deformation amplitude (r = −0.384 and −0.375, respectively, p<0.05) in both groups. Compared to post-PRK eyes, virgin eyes showed longer 1_st applanation time (7.29±0.21 vs. 6.96±0.17 ms, p<0.05) and lower deformation amplitude (1.06±0.07 vs. 1.17±0.08 mm, p<0.05).

Conclusions

CorVis ST demonstrated reliable measurements for CCT, IOP, and 1_st applanation time, as well as relatively reliable measurement for deformation amplitude in both virgin and post-PRK eyes. There were differences in 1_st applanation time and deformation amplitude between virgin and post-PRK eyes, which may reflect corneal biomechanical changes occurring after the surgery in the latter.     

Classification of immature and mature cells of the neutrophil series using morphometrical parameters

Quantitative morphological data of six classes of immature and mature cells of the neutrophil series of the bone marrow of normal persons were used for statistical classification experiments (myeloblasts, promyelocytes, myelocytes, metamyelocytes, bands and segments). On each cell, parameters were measured directly from the image or calculated from the shape of the density histogram or the counting densitogram using a Texture Analysis System (E. Leitz, Wetzlar, Germany). The parameters were analyzed with the interactive statistical pattern recognition system ISPAHAN. One half of the data were used as a learning set and the other half as the test set. The parameters were compared according to their performance in discrimination between the classes, alone and in combinations. Parameters not contributing to an improvement of the discrimination were disregarded. Eleven parameters were selected and used for classification by two different methods: a stepwise and a "one-shot" method. Stepwise classification resulted in a 79% correct classification rate. Most errors occurred between cell classes in neighboring stages of maturation. In 96% of all cases the computer classification was either in accordance with that of the technician or with a cell class of a neighboring maturation stage. One step classification by the computer was in agreement with the technicians in 82% of the cases. For 98% of the cells the computer classification was either in accordance with that of the technician or with a cell class of a neighboring maturation stage. The data set was collected by two technicians, operating independently. Differences in their interpretation of the maturation stage were found by comparing the performance of classifiers based on both cell samples. Since the images of the cells were not available for reexamination, the causes of disagreement in classification between the technicians and between computer and technicians could not be evaluated.     

Reproducibility of power production during sprint arm ergometry

Sprint tests are frequently used to evaluate between-subject differences and can provide a valuable insight into performance capacity. The present study determined the reproducibility of peak and mean power output during upper-body sprints. After familiarization 25 men (mean [+/- SD] age 29 [6] years, body mass 82.8 [12.7] kg and height 1.76 [0.05] m) completed 2 20-second upper-body sprint tests using an adapted cycle ergometer. Mean (+/- SD) values of all power (uncorrected and corrected) measurements achieved during the 2 tests were checked for systematic bias using separate paired t-tests. Test-retest reproducibility was examined using coefficients of variation and single-measure intraclass correlation coefficients, as well as an assessment of the typical (random) error and the 95% limits of agreement. The value of corrected peak power (628 [167] W) was higher (p < 0.05) compared with the uncorrected value (509 [109] W). Values of corrected (465 [95] W) and uncorrected (444 [87] W) mean power were similar (p > 0.05). The mean bias value for all power parameters equated to less than +/-1% of the absolute values of power measured. Intraclass correlation coefficients for all data sets ranged from 0.97 to 0.98. Coefficients of variation for uncorrected and corrected values of peak power were 2.8 and 4.5%, while corresponding values for mean power were 2.9 and 3.2%, respectively. The reproducibility of all power indices was below 5%. The results of this study indicate that both uncorrected and corrected measurements of peak power output and mean power output can be used to assess performance during sprint arm ergometry.     

Small-sample reproducibility estimates: an example using rebreathing measurements

We developed a statistical technique to estimate the reproducibility of a parameter from a population in which only two repeated measurements can be made in a single individual. The following data were analyzed: acetylene cardiac output (Qc), lung tissue volume (Vti), and carbon monoxide diffusing capacity (DLCO) measured by rebreathing techniques in a population of 86 healthy subjects (51 men and 35 women). Each subject was measured twice with a computerized rebreathing system using a test gas of 10% He-0.3% C18O-0.7% C2H2-25% O2-balance N2 while sitting at rest. The estimated coefficients of variation for repeated measurements were 6.8, 10.3, and 5.7% for Qc, Vti, and DLCO, respectively. Chebyshev's inequality was used to estimate the imprecision for a single measurement of these parameters and for averages of two or more repeated values. A single measurement of Qc would be within 14.2% of a "true" mean 90% of the time, whereas an average of three consecutive measurements would be within 8.2% of the true mean 90% of the time. Single measurements of Vti and DLCO were found to be within 21.7 and 12.0%, respectively, of the true mean 90% of the time. When three consecutive measurements are averaged, Vti is within 12.6% and DLCO is within 6.9% of the true mean 90% of the time. We conclude that 1) rebreathing Qc is as reproducible as other measurements of cardiac output, 2) rebreathing measurements of DLCO are as reproducible as those made by the single-breath technique, and 3) an average of two to three measurements of Vti should be made to obtain values with a reasonable degree of precision.     

Implementation of flow cytometry for quality control in four Danish bull studs

A flow cytometric method for simultaneous determination of sperm concentration and viability has recently been developed. In 2001, four Danish bull studs purchased flow cytometers and eight technicians were trained for routine analysis of raw and frozen-thawed semen. After initial training of the technicians, an experiment was carried out to document the precision of the system. The aim was also to assess if flow cytometric determination of sperm concentration could result in a more uniform production of semen doses. Results of this experiment showed high precision in the determination of sperm concentration and coefficients of variation were 3.5 and 2.4% for raw and frozen-thawed semen, respectively. Sperm viability was also assessed with high precision and coefficients of variation were 0.9% for raw semen and 1.7% for frozen-thawed semen. Furthermore, the experiment showed that package of semen doses after flow cytometric determination of sperm concentration in the raw semen results in a significantly smaller variation in the number of sperm per dose. In the second experiment, frozen semen was exchanged between the participating studs and were analysed by flow cytometry as well as by microscopic assessments by the eight technicians. Results show that the average correlation between technicians were 0.38 for motility assessments while flow cytometric agreement between technicians was significantly higher (average correlation was 0.86 for sperm viability and 0.92 for sperm concentration). The experiment also showed very high agreements between assessments within lab technician (correlations r=0.98 (sperm concentration) and r=0.99 (sperm viability)). Experiment 3 revealed that straws from the same batch varies in both concentration and viability. It is concluded that flow cytometric determination of sperm concentration and viability can be used to improve semen assessment by AI studs and result in a better quality control.     

Non-patient related variables affecting levels of vascular endothelial growth factor in urine biospecimens

Vascular endothelial growth factor (VEGF) is an angiogenic protein proposed to be an important biomarker for the prediction of tumour growth and disease progression. Recent studies suggest that VEGF measurements in biospecimens, including urine, may have predictive value across a range of cancers. However, the reproducibility and reliability of urinary VEGF measurements have not been determined. We collected urine samples from patients receiving radiation treatment for glioblastoma multiforme (GBM) and examined the effects of five variables on measured VEGF levels using an ELISA assay. To quantify the factors affecting the precision of the assay, two variables were examined: the variation between ELISA kits with different lot numbers and the variation between different technicians. Three variables were tested for their effects on measured VEGF concentration: the time the specimen spent at room temperature prior to assay, the addition of protease inhibitors prior to specimen storage and the alteration of urinary pH. This study found that VEGF levels were consistent across three different ELISA kit lot numbers. However, significant variation was observed between results obtained by different technicians. VEGF concentrations were dependent on time at room temperature before measurement, with higher values observed 3-7 hrs after removal from the freezer. No significant difference was observed in VEGF levels with the addition of protease inhibitors, and alteration of urinary pH did not significantly affect VEGF measurements. In conclusion, this determination of the conditions necessary to reliably measure urinary VEGF levels will be useful for future studies related to protein biomarkers and disease progression.     

A comparison of two techniques for measuring and crossdating tree rings

Crossdating is the core principle of dendrochronology. Our study compared two techniques for measuring and crossdating tree rings using Juniperus virginiana L. (eastern redcedar) as a case study. We used a pseudo 2 × 2 study design comparing the traditional skeleton plot/sliding measuring stage technique to a semi-automatic image analysis program across two technicians. Crossdating was evaluated in COFECHA. Raw measurements of total, earlywood, and latewood widths from the two methods were analyzed using the Verify for Windows program, ANOVA, and correlation matrices. Total ring width and earlywood width were well correlated between techniques and technicians but questionable ring boundaries from image analysis program should be checked under a stereoscope. Juniperus virginiana latewood widths were significantly different between techniques and technicians; therefore, we do not recommend combining latewood measurements from species with limited latewood variability for dendrochronological analysis. A standard definition of the earlywood-latewood boundary that can be replicated across technicians is needed to combine latewood measurements from the sliding measuring stage and image analysis systems.     

Inter Subject Variability and Reproducibility of Diffusion Tensor Imaging within and between Different Imaging Sessions

The aim of these studies was to provide reference data on intersubject variability and reproducibility of diffusion tensor imaging. Healthy volunteers underwent imaging on two occasions using the same 3T Siemens Verio magnetic resonance scanner. At each session two identical diffusion tensor sequences were obtained along with standard structural imaging. Fractional anisotropy, apparent diffusion coefficient, axial and radial diffusivity maps were created and regions of interest applied in normalised space. The baseline data from all 26 volunteers were used to calculate the intersubject variability, while within session and between session reproducibility were calculated from all the available data. The reproducibility of measurements were used to calculate the overall and within session 95% prediction interval for zero change. The within and between session reproducibility data were lower than the values for intersubject variability, and were different across the brain. The regional mean (range) coefficient of variation figures for within session reproducibility were 2.1 (0.9–5.5%), 1.2 (0.4–3.9%), 1.2 (0.4–3.8%) and 1.8 (0.4–4.3%) for fractional anisotropy, apparent diffusion coefficient, axial and radial diffusivity, and were lower than between session reproducibility measurements (2.4 (1.1–5.9%), 1.9 (0.7–5.7%), 1.7 (0.7–4.7%) and 2.4 (0.9–5.8%); p<0.001). The calculated overall and within session 95% prediction intervals for zero change were similar. This study provides additional reference data concerning intersubject variability and reproducibility of diffusion tensor imaging conducted within the same imaging session and different imaging sessions. These data can be utilised in interventional studies to quantify change within a single imaging session, or to assess the significance of change in longitudinal studies of brain injury and disease.     

Comparison of Inter Subject Variability and Reproducibility of Whole Brain Proton Spectroscopy

The aim of these studies was to provide reference data on intersubject variability and reproducibility of metabolite ratios for Choline/Creatine (Cho/Cr), N-acetyl aspartate/Choline (NAA/Cho) and N-acetyl aspartate/Creatine (NAA/Cr), and individual signal-intensity normalised metabolite concentrations of NAA, Cho and Cr. Healthy volunteers underwent imaging on two occasions using the same 3T Siemens Verio magnetic resonance scanner. At each session two identical Metabolic Imaging and Data Acquisition Software (MIDAS) sequences were obtained along with standard structural imaging. Metabolite maps were created and regions of interest applied in normalised space. The baseline data from all 32 volunteers were used to calculate the intersubject variability, while within session and between session reproducibility were calculated from all the available data. The reproducibility of measurements were used to calculate the overall and within session 95% prediction interval for zero change. The within and between session reproducibility data were lower than the values for intersubject variability, and were variable across the different brain regions. The within and between session reproducibility measurements were similar for Cho/Cr, NAA/Choline, Cho and Cr (11.8%, 11.4%, 14.3 and 10.6% vs. 11.9%, 11.4%, 13.5% and 10.5% respectively), but for NAA/Creatine and NAA between session reproducibility was lower (9.3% and 9.1% vs. 10.1% and 9.9%; p <0.05). This study provides additional reference data that can be utilised in interventional studies to quantify change within a single imaging session, or to assess the significance of change in longitudinal studies of brain injury and disease.     

Dynamic integration of reward and stimulus information in perceptual decision-making

In perceptual decision-making, ideal decision-makers should bias their choices toward alternatives associated with larger rewards, and the extent of the bias should decrease as stimulus sensitivity increases. When responses must be made at different times after stimulus onset, stimulus sensitivity grows with time from zero to a final asymptotic level. Are decision makers able to produce responses that are more biased if they are made soon after stimulus onset, but less biased if they are made after more evidence has been accumulated? If so, how close to optimal can they come in doing this, and how might their performance be achieved mechanistically? We report an experiment in which the payoff for each alternative is indicated before stimulus onset. Processing time is controlled by a “go” cue occurring at different times post stimulus onset, requiring a response within msec. Reward bias does start high when processing time is short and decreases as sensitivity increases, leveling off at a non-zero value. However, the degree of bias is sub-optimal for shorter processing times. We present a mechanistic account of participants'' performance within the framework of the leaky competing accumulator model [1], in which accumulators for each alternative accumulate noisy information subject to leakage and mutual inhibition. The leveling off of accuracy is attributed to mutual inhibition between the accumulators, allowing the accumulator that gathers the most evidence early in a trial to suppress the alternative. Three ways reward might affect decision making in this framework are considered. One of the three, in which reward affects the starting point of the evidence accumulation process, is consistent with the qualitative pattern of the observed reward bias effect, while the other two are not. Incorporating this assumption into the leaky competing accumulator model, we are able to provide close quantitative fits to individual participant data.     

Technical bias and the reproducibility crisis: The problem of systemic errors resulting from artefacts of equipment,methods or dataset has been underappreciated

Technical bias can cause lack of reproducibility. While harder to identify than other bias, it can cause consistent systemic errors in experimental data and analysis. Subject Categories: Methods & Resources, S&S: Ethics

The problem of reproducibility of experimental results has been festering for some years and prompted a great deal of research and analysis to explain why so many studies in biomedical research cannot be validated. Among these, the role played by technical bias in yielding erroneous results has been largely underappreciated in contrast to better‐known sources of error, such as inadequate sample sizes, selective reporting or publication pressure. But its significance has come to the fore after recent studies raised questions over earlier findings that were shown to have been warped by technical bias.

Technical bias often occurs alongside other forms of bias and the effects can be additive as well as overlapping.

Technical bias has nothing directly to do with human psychology or deficient analysis, but results from artefacts of equipment, reagents and laboratory methods, as well as lack of standard protocols. As a result, it has contributed to the reproducibility crisis both between (inter) and within (intra) institutions or even individual laboratories. Technical bias often occurs alongside other forms of bias and the effects can be additive as well as overlapping. Some of these apply more to particular fields than others, with sociological research tending to be plagued by different biases than say physics or chemistry. Overall, one could identify six broad categories alongside technical bias.The one most closely related to technical bias is statistical bias, sometimes called sampling error, which arises when the data are not representative of the wider set from which the results are expected. Either the sampling size for an experiment or survey is too small or it is skewed towards a subsector of the whole. The latter can often distort surveys purely because people who agree to take part often differ in some ways from those who decline to do so.Then, there is measurement bias caused by systematic errors in the process of collecting or generating data. A simple example would be measuring height without asking people to take off their shoes. The error would vary between individuals but the apparent average and mean would be too high.Third is confirmation bias, often defined as an unconscious tendency to seek out, recall or notice data that confirms preconceived opinions or hypotheses. This has been cited in the context of the COVID‐19 pandemic where aspects such as hospitalization levels or mutation rates are taken to confirm worst fears when they may be just normal artefacts of a respiratory virus.Fourth is design bias, where misleading results can arise out of the way an experiment or investigation is constructed. This has overlap with some of the other biases since design involves selection of samples, generation of data and the methods used.Fifth is publication bias: the tendency to pursue research and to report results that have a positive outcome, reinforced by journals'' preference for such papers. This is closely related to positive outcome bias and can lead to an overall misleading and rosy picture of research in a field. Egregious examples include the infamous Andrew Wakefield paper falsely connecting the MMR vaccine with autism, which may well not have been published without this tendency towards publicity‐grabbing headlines.Finally, there is hindsight bias, a common human affliction that is the tendency to regard events as having been predictable only after they have occurred. This can lead researchers to subconscious revisionism, for example raising the average prior probability of a given outcome after that has come to pass.

… there is hindsight bias, a common human affliction that is the tendency to regard events as having been predictable only after they have occurred.

     

Are exhaled nitric oxide measurements using the portable NIOX MINO repeatable?

Background

Exhaled nitric oxide is a non-invasive marker of airway inflammation and a portable analyser, the NIOX MINO (Aerocrine AB, Solna, Sweden), is now available. This study aimed to assess the reproducibility of the NIOX MINO measurements across age, sex and lung function for both absolute and categorical exhaled nitric oxide values in two distinct groups of children and teenagers.

Methods

Paired exhaled nitric oxide readings were obtained from 494 teenagers, aged 16-18 years, enrolled in an unselected birth cohort and 65 young people, aged 6-17 years, with asthma enrolled in an interventional asthma management study.

Results

The birth cohort participants showed a high degree of variability between first and second exhaled nitric oxide readings (mean intra-participant difference 1.37 ppb, 95% limits of agreement -7.61 to 10.34 ppb), although there was very close agreement when values were categorised as low, normal, intermediate or high (kappa = 0.907, p < 0.001). Similar findings were seen in subgroup analyses by sex, lung function and asthma status. Similar findings were seen in the interventional study participants.

Conclusions

The reproducibility of exhaled nitric oxide is poor for absolute values but acceptable when values are categorised as low, normal, intermediate or high in children and teenagers. One measurement is therefore sufficient when using categorical exhaled nitric oxide values to direct asthma management but a mean of at least two measurements is required for absolute values.     

An environmental signature for 323 microbial genomes based on codon adaptation indices

Background

Codon adaptation indices (CAIs) represent an evolutionary strategy to modulate gene expression and have widely been used to predict potentially highly expressed genes within microbial genomes. Here, we evaluate and compare two very different methods for estimating CAI values, one corresponding to translational codon usage bias and the second obtained mathematically by searching for the most dominant codon bias.

Results

The level of correlation between these two CAI methods is a simple and intuitive measure of the degree of translational bias in an organism, and from this we confirm that fast replicating bacteria are more likely to have a dominant translational codon usage bias than are slow replicating bacteria, and that this translational codon usage bias may be used for prediction of highly expressed genes. By analyzing more than 300 bacterial genomes, as well as five fungal genomes, we show that codon usage preference provides an environmental signature by which it is possible to group bacteria according to their lifestyle, for instance soil bacteria and soil symbionts, spore formers, enteric bacteria, aquatic bacteria, and intercellular and extracellular pathogens.

Conclusion

The results and the approach described here may be used to acquire new knowledge regarding species lifestyle and to elucidate relationships between organisms that are far apart evolutionarily.     

Evaluation of bull fertility in dairy and beef cattle using cow field data

A successful outcome to a given service is a combination of both male and female fertility. Despite this, most national evaluations for fertility are generally confined to female fertility with evaluations for male fertility commonly undertaken by individual breeding organisations and generally not made public. The objective of this study was to define a pertinent male fertility trait for seasonal calving production systems, and to develop a multiple regression mixed model that may be used to evaluate male fertility at a national level. The data included in the study after editing consisted of 361,412 artificial inseminations from 206,683 cow-lactations (134,911 cows) in 2,843 commercial dairy and beef herds. Fixed effects associated with whether a successful pregnancy ensued (pregnant = 1) or not (pregnant = 0) from a given service were year by month of service, day of the week, days since calving, cow parity, level of calving difficulty experienced, whether or not the previous calving was associated with perinatal mortality, and age of the service bull at the date of insemination. Non-additive genetic effects such as heterosis and recombination loss as well as inbreeding level of the service bull, dam or mating were not associated with a successful pregnancy; there was no difference in pregnancy rate between fresh or frozen semen. Random effects included in the model were the additive genetic effect of the cow, as well as a within lactation and across lactation permanent environmental effect of the cow; pedigree group effects based on cow breed were also included via the relationship matrix. Temporal differences in the AI technician and service bull were also included as random effects. A difference in five percentage units in male fertility was evident between the average effects of different dairy and beef breeds. The correlation between raw pregnancy rates for bulls with more than 100 services (n = 431) and service bull solutions from the mixed model analysis was 0.66. The correlation between the raw pregnancy rates of 288 technicians with more than 100 services and their respective solutions from the mixed model was 0.35. These low to moderate correlations suggest considerable re-ranking among both service bulls and technicians and suggest possibly a benefit of using a statistical model to better estimate the performance of both service bulls and technicians.     

Measurement errors and regression to the mean cannot explain bias in average Ellenberg indicator values

Smart & Scott (2004, this is sue) criticized our paper (Wamelink et al. 2002) about the bias in average Ellenberg indicator values. Their main criticism concerns the method we used, regression analysis. They state the bias can be mimicked by the construction of an artificial data set and that regression analysis is not a suited tool to investigate underlying phenomena. Moreover they claim that the present bias is caused by the distribution of Ellenberg indicator values between syntaxa, instead of a bias in average Ellenberg indicator values per species. We show that their criticism of the use of regression analysis does not hold. We selected average Ellenberg values per vegetation group for several pH classes and applied an F‐test to determine whether or not the vegetation groups within each pH class differed significantly from each other. This was the case for all tested classes (P < 0.001). Moreover we simulated an artificial data set, of which the F‐test for varying measurement error could not explain the magnitude of the F‐value we found earlier. This indicates that the bias we found in average Ellenberg indicator values cannot be explained by measurement errors or by regression to the mean. In the end, Smart & Scott, as we did, come to the conclusion that there is a bias present and that separate regression lines per vegetation type are necessary, but the debate remains open on whether or not this is caused by the bias in Ellenberg indicator values per species.