Is cross-validation better than resubstitution for ranking genes?

MOTIVATION: Ranking gene feature sets is a key issue for both phenotype classification, for instance, tumor classification in a DNA microarray experiment, and prediction in the context of genetic regulatory networks. Two broad methods are available to estimate the error (misclassification rate) of a classifier. Resubstitution fits a single classifier to the data, and applies this classifier in turn to each data observation. Cross-validation (in leave-one-out form) removes each observation in turn, constructs the classifier, and then computes whether this leave-one-out classifier correctly classifies the deleted observation. Resubstitution typically underestimates classifier error, severely so in many cases. Cross-validation has the advantage of producing an effectively unbiased error estimate, but the estimate is highly variable. In many applications it is not the misclassification rate per se that is of interest, but rather the construction of gene sets that have the potential to classify or predict. Hence, one needs to rank feature sets based on their performance. RESULTS: A model-based approach is used to compare the ranking performances of resubstitution and cross-validation for classification based on real-valued feature sets and for prediction in the context of probabilistic Boolean networks (PBNs). For classification, a Gaussian model is considered, along with classification via linear discriminant analysis and the 3-nearest-neighbor classification rule. Prediction is examined in the steady-distribution of a PBN. Three metrics are proposed to compare feature-set ranking based on error estimation with ranking based on the true error, which is known owing to the model-based approach. In all cases, resubstitution is competitive with cross-validation relative to ranking accuracy. This is in addition to the enormous savings in computation time afforded by resubstitution.     

Does loss of heterozygosity in critical genome regions predict a local relapse in patients after laryngectomy?

BACKGROUND: Patients, who had an upper aerodigestive tract malignancy, have a high incidence of succeeding tumor development. This has been attributed to the role of "field cancerization" in carcinogenesis. The aim of this study was analysis of loss of heterozygosity (LOH) in the regions frequently lost during the course of head and neck squamous cell carcinomas (HNSCC), especially at early stages, which could answer the clinicians' question, if LOH analysis has any "predictive" value in relation to tumor occurrence. MATERIAL AND METHODS: Sixty-five larynx cancer patients were examined for loss of heterozygosity on 3p, 7q, 8p, 9p and 18q chromosomal arms with the use of 12 microsatellite markers. The material from a single patient consisted of blood, tumor, safe margin and one or two clinically unchanged mucosal samples. During follow up, the material from brush specimens (14 patients) as well as laryngeal swabs (4 patients) was also examined. RESULTS: The highest frequency of LOH was detected for marker D3S1234 in tumor tissues (29%). Analysis of margin samples (b) revealed low LOH frequencies (2-5%) and complete retention of heterozygosity for markers: D3S1234, D7S486, D8S261, D8S264, D9S171 and D18S46. Similarly, for normal appearing mucosa from upper part of larynx (c) frequencies of LOH were low (2-6%), with the complete retention of heterozygosity for markers: D3S1284, D3S1304, D3S1234, D8S264 and D9S1870. We did not detect any LOH in the material of normal appearing mucosa from tracheostoma region (d). During follow up, LOH was detected for eight markers, with the highest incidence for markers D18S46 (six cases), D7S486 (four cases) and D3S1300 (three cases). CONCLUSIONS: The data, obtained during this investigation, did not reveal the predictive value of LOH with respect to local relapse occurrence in laryngeal cancer patients. However, time of follow up did not reach 5 years, so that further clinical monitoring should be conducted.     

Which evaluation criteria of the short-term efficacy can better reflect the long-term outcomes for patients with nasopharyngeal carcinoma?

PurposeTo compare the consistency of one-dimensional Response Evaluation Criteria in Solid Tumors (1D-RECIST), two-dimensional WHO criteria (2D-WHO), and three-dimensional (3D) measurement for therapeutic response assessment of nasopharyngeal carcinoma (NPC).Materials and methodsRetrospective data of 288 newly diagnosed NPC patients were reviewed. Tumor size was assessed on magnetic resonance imaging (MRI) according to the 1D-RECIST, 2D-WHO, and 3D measurement criteria. Agreement between tumor responses was assessed using unweighted k statistics. The receiver operating characteristic (ROC) curve was used to determine the optimal cut-off point of the PTV. The Kaplan–Meier method and Cox regression were used for the survival analysis.ResultsThe optimal cut-off point of the PTV for progression-free survival (PFS) was 29.6%. Agreement with PTV measurement was better for 1D measurement than for 2D and 3D measurements (kappa values of 0.646, 0.537, and 0.577 for 1D, 2D, and 3D measurements, respectively; P < 0.05). The area under the curve of the 1D measurement (AUC=0.596) was similar to that of the PTV measurement (AUC=0.621). Compared with 2D and 3D measurements, 1D measurement is superior for predicting prognosis in NPC (C-index of 0.672, 0.663, and 0.646 were for 1D, 2D, and 3D measurements, respectively; P < 0.005). Survival analysis showed that patients with non-responders had worse prognosis (P < 0.05).ConclusionsThe 1D measurement more closely agreed with the PTV measurement than the 2D and 3D measurements for predicting therapeutic responses in NPC. Therefore, we recommend using the less time-consuming 1D-RECIST criteria in routine clinical practice.     

Does transforming growth factor-beta1 predict for radiation-induced pneumonitis in patients treated for lung cancer?

The purpose of the study was to reassess the utility of transforming growth factor-beta-1 (TGF-beta1) together with dosimetric and tumor parameters as a predictor for radiation pneumonitis (RP). Of the 121 patients studied, 32 (26.4%) developed grade > or =1 RP, and 27 (22.3%) developed grade > or =2 RP. For the endpoint of grade > or =1 RP, those with V30>30% and an end-RT/baseline TGF-beta1 ratio> or =1 had a significantly higher incidence of RP than did those with V30>30% and an end-RT/baseline TGF-beta1 ratio<1. For most other patient groups, there were no clear associations between TGF-beta1 values and rates of RP. These findings suggest that TGF-beta1 is generally not predictive for RP except for the group of patients with a high V30.     

Grooming in Barbary macaques: better to give than to receive?

It is well established that grooming underpins sociality in group-living primates, and a number of studies have documented the stress-reducing effects of being groomed. In this study, we quantified grooming behaviour and physiological stress (assessed by faecal glucocorticoid analysis) in free-ranging Barbary macaques, Macaca sylvanus. Our results indicate that it is the giving rather than the receiving of grooming that is associated with lower stress levels. These findings shed important new light on the benefits of this key behaviour in primate social life.     

Two-dimensional gel electrophoresis; better than a poke in the ICAT?

To date, the most widely used technology for conducting proteomic studies has been two-dimensional gel electrophoresis (2DGE), but this approach does have drawbacks. Isotope-coded affinity tagging (ICAT) is starting to challenge 2DGE as a new proteomic tool for the analysis of proteins in complex biological specimens. An appraisal of these two methodologies reveals that neither ICAT nor 2DGE provide comprehensive coverage on a proteome-wide scale.     

Distribution and fate of DCX/PSA-NCAM expressing cells in the adult mammalian cortex: A local reservoir for adult cortical neuroplasticity?

The expression of early developmental markers such as doublecortin (DCX) and the polysialylated-neural cell adhesion molecule (PSA-NCAM) has been used to identify immature neurons within canonical neurogenic niches. Additionally, DCX/PSA-NCAM+ immature neurons reside in cortical layer II of the paleocortex and in the paleo- and entorhinal cortex of mice and rats, respectively. These cells are also found in the neocortex of guinea pigs, rabbits, some afrotherian mammals, cats, dogs, non-human primates, and humans. The population of cortical DCX/PSA-NCAM+ immature neurons is generated prenatally as conclusively demonstrated in mice, rats, and guinea pigs. Thus, the majority of these cells do not appear to be the product of adult proliferative events. The immature neurons in cortical layer II are most abundant in the cortices of young individuals, while very few DCX/PSA-NCAM + cortical neurons can be detected in aged mammals. Maturation of DCX/PSA-NCAM+ cells into glutamatergic and GABAergic neurons has been proposed as an explanation for the age-dependent reduction in their population over time. In this review, we compile the recent information regarding the age-related decrease in the number of cortical DCX/PSA-NCAM+ neurons. We compare the distribution and fates of DCX/PSA-NCAM + neurons among mammalian species and speculate their impact on cognitive function. To respond to the diversity of adult neurogenesis research produced over the last number of decades, we close this review by discussing the use and precision of the term “adult non-canonical neurogenesis.”     

Are there risk factors for acute renal failure in adult patients using deoxycholate amphotericin B?

BackgroundDeoxycholate amphotericin B (DAB) is a nephrotoxic drug and the incidence of acute kidney injury (AKI) is high.AimsThe aim of this study was to describe the incidence of AKI in patients under DAB therapy and determine risk factor to predict the AKI.MethodsThe data of this retrospective study included previously hospitalized patients treated with intravenous DAB for at least five days. Clinical and laboratorial data were evaluated and AKI was classified in stages using Acute Kidney Injury Network criteria. Univariated test followed by a multivariable analysis was performed to determine risk factor and Kaplan–Meier survival estimates were calculated to evaluate the role of AKI in the outcome.ResultsOne hundred six patients were included in the final analysis. AKI occurred in 51.9% and dialysis was necessary in 4.7%. The occurrence of AKI was not associated with any risk factor. The mortality of the patients was neither associated with AKI nor with dialysis. Other nephrotoxic drugs were not risk factors for AKI.ConclusionsThe incidence of AKI in patients using DAB is high and we cannot predict the chance of AKI using clinical or laboratorial data.     

Does a change in sleep timing increase testosterone in young adult men?

Men are more evening oriented than women. These sex difference might be a result of gonadal hormones, especially of testosterone. Here, we tested for a causal relationship by shifting the sleep timing of young men to one hour later and assessed the influence on saliva testosterone. 50 men participated in our study (31 experimental, 19 control group). Sleep timing was measured with actigraphy. One week served as baseline and afterwards, participants were randomly assigned to either the control or the experimental group. In the experimental group, men were asked to go to bed about one hour later and to sleep one hour longer in the morning. Participants in the control group experienced no significant changes in sleep timing. In the experimental group, participants went to bed later and woke up later; therefore, mid-point of sleep was shifted to a later time and sleep duration decreased, because he men went to bed about one hour later but woke up only about 30 min later. The control group experienced a decline in testosterone, but the experimental group not. Changes in mid-point of sleep between baseline week and experimental week were correlated with the change in testosterone (r = 0.285).     

Why is the number of days required for induction of adult diapause in the linden bug Pyrrhocoris apterus fewer in the larval than in the adult stage?

Adult females of Pyrrhocoris apterus, programmed for diapause by short-day (SD) photoperiod and those programmed for reproduction by long-day (LD) retain photoperiodic information in continuous darkness (DD) until death. However, if the interruption of SD by DD is made in the course of diapause programming in adults, then the incidence of diapause depends on the number of SD cycles received before DD, with no evidence that the photoperiodic clock is free-running at DD to complete diapause induction. These results indicate that the photoperiodic clock is stopped after transfer to DD and the information accumulated before transfer to DD is maintained. Diapause programming in the adult stage requires 9–10 SD cycles to induce diapause in 80% of individuals. However, if the diapause programming starts after ecdysis of LD-larvae to the last instar, only 3 SD cycles before transfer to DD are required for diapause in 80% of individuals. Surprisingly, if the newly ecdysed last instar LD-larvae, sensitive to photoperiod, are transferred to DD (thus they did not experience any SD), diapause occurs in 40% of the individuals. Thus, diapause ‘information’ is present in LD-larvae and is responsible for a lower number of SD required for diapause induction in the larval than in the adult stage.     

ITS1: a DNA barcode better than ITS2 in eukaryotes?

A DNA barcode is a short piece of DNA sequence used for species determination and discovery. The internal transcribed spacer (ITS/ITS2) region has been proposed as the standard DNA barcode for fungi and seed plants and has been widely used in DNA barcoding analyses for other biological groups, for example algae, protists and animals. The ITS region consists of both ITS1 and ITS2 regions. Here, a large‐scale meta‐analysis was carried out to compare ITS1 and ITS2 from three aspects: PCR amplification, DNA sequencing and species discrimination, in terms of the presence of DNA barcoding gaps, species discrimination efficiency, sequence length distribution, GC content distribution and primer universality. In total, 85 345 sequence pairs in 10 major groups of eukaryotes, including ascomycetes, basidiomycetes, liverworts, mosses, ferns, gymnosperms, monocotyledons, eudicotyledons, insects and fishes, covering 611 families, 3694 genera, and 19 060 species, were analysed. Using similarity‐based methods, we calculated species discrimination efficiencies for ITS1 and ITS2 in all major groups, families and genera. Using Fisher's exact test, we found that ITS1 has significantly higher efficiencies than ITS2 in 17 of the 47 families and 20 of the 49 genera, which are sample‐rich. By in silico PCR amplification evaluation, primer universality of the extensively applied ITS1 primers was found superior to that of ITS2 primers. Additionally, shorter length of amplification product and lower GC content was discovered to be two other advantages of ITS1 for sequencing. In summary, ITS1 represents a better DNA barcode than ITS2 for eukaryotic species.     

Does prophylactic thyroxine treatment after operation for non-toxic goitre influence thyroid size?

In order to evaluate the influence of thyroxine treatment on thyroid volume after thyroidectomy for non-toxic goitre 110 consecutive patients were randomised to receive thyroxine (150 micrograms daily) or no treatment three months after operation. Thyroid volume determined by ultrasonography did not differ significantly between the two groups one year after operation. Nevertheless, a similar decrease in volume was seen from three to 12 months postoperatively in the thyroxine treated and no treatment groups (median 18 (range 8-70) ml to 16 (range 7-57) ml, and median 20 (range 9-72) ml to 17 (range 8-58) ml, respectively). Three patients (two given thyroxine) had recurrence of goitre within the observation period of one year. In this series thyroid volume decreased during the first postoperative year independently of thyroxine treatment. Hence it seems questionable whether postoperative thyroid function studies can identify patients at risk of recurrence of goitre after operation for non-toxic goitre and whether routine postoperative treatment with thyroxine is justifiable.