An averaging strategy to reduce variability in target-decoy estimates of false discovery rate

Decoy database search with target-decoy competition (TDC) provides an intuitive, easy-to-implement method for estimating the false discovery rate (FDR) associated with spectrum identifications from shotgun proteomics data. However, the procedure can yield different results for a fixed dataset analyzed with different decoy databases, and this decoy-induced variability is particularly problematic for smaller FDR thresholds, datasets or databases. In such cases, the nominal FDR might be 1% but the true proportion of false discoveries might be 10%. The averaged TDC protocol combats this problem by exploiting multiple independently shuffled decoy databases to provide an FDR estimate with reduced variability. We provide a tutorial introduction to aTDC, describe an improved variant of the protocol that offers increased statistical power, and discuss how to deploy aTDC in practice using the Crux software toolkit.

Can redirected aggression explain interspecific attacks by Australian magpies on other birds?

Attacks by participants of conflicts against a third party are referred to as redirected aggression. Usually the third party is a conspecific — few documented cases of redirected aggression against other species exist. The Australian magpie (Gymnorhina tibicen), however, often attacks other species; the reasons for attacks are largely unknown. Some attacks occur after territorial disputes with conspecifics, suggesting that attacks are the result of redirected aggression. We subjected eight Australian magpie groups to simulated territorial intrusions. In one treatment an Australian magpie and rock dove (Columba livia) were presented in cages next to one another (5 m apart) on the territory for 30 min; the Australian magpie decoy was then covered and aggressive responses toward the rock dove by residents were recorded for a further 30 min (Treatment 1). Two additional treatments were presented in an identical manner on each territory where both decoys were either Australian magpies (Treatment 2) or rock doves (Treatment 3). We predicted that if Australian magpies regularly redirect aggression onto benign species after conspecific territorial intrusions, attack rates on the rock dove decoy in Treatment 1 would be higher than attack rates on the rock dove decoy in Treatment 3. Residents were seldom recorded close (<1 m) and not seen attacking rock dove decoys during tests. In contrast, Australian magpie decoys were often approached and attacked by residents. After a decoy was covered (following the first 30 min phase), residents spent little time in close proximity (on cage, <0.3 m, or 0.3-1 m) to the rock dove decoy in either Treatment 1 or Treatment 3; in contrast, residents were often recorded close to the Australian magpie decoy in Treatment 2. We found no evidence that Australian magpies redirect aggression onto other birds after territorial intrusions. The true proportion of territorial disputes leading to redirected attacks may be small, or only occur under highly specific contexts.

Enhancing Local Health Department Disaster Response Capacity with Rapid Community Needs Assessments: Validation of a Computerized Program for Binary Attribute Cluster Sampling

Introduction:Local health departments are among the first agencies to respond to disasters or other mass emergencies. However, they often lack the ability to handle large-scale events. Plans including locally developed and deployed tools may enhance local response. Simplified cluster sampling methods can be useful in assessing community needs after a sudden-onset, short duration event.Methods:Using an adaptation of the methodology used by the World Health Organization Expanded Programme on Immunization (EPI), a Microsoft Access-based application for two-stage cluster sampling of residential addresses in Louisville/Jefferson County Metro, Kentucky was developed. The sampling frame was derived from geographically referenced data on residential addresses and political districts available through the Louisville/Jefferson County Information Consortium (LOJIC). The program randomly selected 30 clusters, defined as election precincts, from within the area of interest, and then, randomly selected 10 residential addresses from each cluster.The program, called the Rapid Assessment Tools Package (RATP), was tested in terms of accuracy and precision using data on a dichotomous characteristic of residential addresses available from the local tax assessor database. A series of 30 samples were produced and analyzed with respect to their precision and accuracy in estimating the prevalence of the study attribute. Point estimates with 95% confidence intervals were calculated by determining the proportion of the study attribute values in each of the samples and compared with the population proportion. To estimate the design effect, corresponding simple random samples of 300 addresses were taken after each of the 30 cluster samples.Results:The sample proportion fell within ±10 absolute percentage points of the true proportion in 80% of the samples. In 93.3% of the samples, the point estimate fell within ±12.5%, and 96.7% fell within ±15%. All of the point estimates fell within ±20% of the true proportion. Estimates of the design effect ranged from 0.926 to 1.436 (mean = 1.157, median = 1.170) for the 30 samples.Conclusions:Although prospective evaluation of its performance in field trials or a real emergency is required to confirm its utility, this study suggests that the Rapid Assessment Tools Package, a locally designed and deployed tool, may provide populationbased estimates of community needs or the extent of event-related consequences that are precise enough to serve as the basis for the initial post-event decisions regarding relief efforts.

The Prevention of Coronary Artery Aneurysm in Kawasaki Disease: A Meta-analysis on the Efficacy of Aspirin and Immunoglobulin Treatment

Objective. Varying observations have been made concerning the use of aspirin (ASA) and/or intravenous immunoglobulin (IVIG) in the prevention of coronary artery aneurysm (CAA) in children with Kawasaki disease. A meta-analysis of published articles on the subject was conducted to evaluate the reported efficacy of these therapies. Methods. All published studies in all languages from 1967 through 1993 obtained from MEDLINE and EMBASE were considered, and a defined set of inclusion and exclusion criteria selected the studies for analysis. These studies were grouped based on whether the children in the studies received: (1) ASA alone, (2) low IVIG (≤1 g/kg) and ASA, (3) high IVIG (&gt;1 g/kg) and ASA, (4) single IVIG (&gt;1 g/kg) and ASA, (5) high IVIG and low ASA (≤80 mg/kg), or (6) high IVIG and high ASA (&gt; 80mg/kg). Studies that satisfied the test for homogeneity were subjected to further analysis. The best estimate of the true proportion of CAA as well as the 95% confidence interval for each group were calculated at 30 and 60 days. Hypothesis testing was conducted to determine the statistical significance of the calculated difference in each compared treatment group. Results. The best estimate of true proportion of CAA and the 95% confidence interval in each group at 30 and 60 days were: (1) ASA group, 30 days, 22.8% (20.6%, 25%); 60 days, 17.1% (13.6%, 20.7%); (2) low-IVIG group, 30 days, 17.3% (14.3%, 20.2%); 60 days, 11.1% (8.7%, 13.6%); (3) high-IVIG group, 30 days, 10.3% (8.3%, 12.3%); 60 days, 4.4% (2.8%, 6%); (4) single-IVIG group, 30 days, 2.3% (0.5%, 4.2%); 60 days, 2.4% (0.5%, 4.2%); (5) high-IVIG-low-ASA group, 30 days, 13% (9%, 17%); 60 days, 4.8% (2.3%, 7.4%); and (6) high-IVIG-high-ASA group, 30 days, 9.1% (6.9%, 11.4%); 60 days, 4% (2%, 6.1%). Conclusion. The incidence of CAA both at 30 and 60 days was significantly lower in low-IVIG than in ASA and in high-IVIG than in low-IVIG groups. Also, the incidence was lower in the single-IVIG than in the high-IVIG group, but this was noted at 30 days and not at 60 days. There was no statistically significant difference in the incidence of CAA both at 30 and 60 days between the high-IWIG-low-ASA and high-IVIG-high-ASA groups.

P. R. Newman and the Durham Protestations

P. R. Newman has done a useful service in tabulating the returns in the Durham Protestations (Recusant History [October 1979], pp. 148-52). Unfortunately, he has omitted the sixty-eight names of persons refusing to sign in the parish of Lanchester, a parish which on the showing of the Protestation itself had an 18% recusancy among its adult males. The inclusion of these would raise Mr Newman's figure of 452 men refusing the Protestation (excluding known non-Catholics) to 520, and the percentage in the county to just under 3.2. An even more accurate percentage might be obtained if the affirmative signatures from Bishop Auckland, a parish which failed to list the refusers, were subtracted from the total. If that were done, the percentage would rise to just over 3.2. Also it may be reasonably assumed that the percentage would be higher still if the whole adult population were taken into account. For, as the lists of prosecutions and more general indications suggest, women were more faithful and persevering adherents of the old religion than men. However, the two-to-one ratio which might be inferred from an examination of the presentments to the Sessions and the church courts probably overstates the true proportion. For this two reasons may be adduced. First, in the early Elizabethan years, when many were uncertain as to where they stood, numerous waverers allowed or perhaps encouraged their wives to take a stand for Catholicism while they themselves made at least an outward show of conformity when that was deemed necessary for the preservation of their families and possessions. Secondly, churchwardens, when pressed for presentments, apparently found that to name widows and spinsters along with vagrants gave rise to less local trouble than the naming of their fellows. Even so, though the proportion of women recusants may have been rather less than the prosecutions suggest, there can be little doubt that they outnumbered men sufficiently for one to speculate that the Catholic adult population may well have been of the order of 3.5%.

Notes upon the Incidence of Tuberculosis in Asylums

It is the prevalent opinion that phthisis is the scourge of our English lunatic asylums, and that these institutions are, literally speaking, hotbeds for the growth and distribution of the tubercle bacillus. In the Irish asylums, where the death-rate from phthisis alone amounts to nearly 30 per cent. of all causes of death, there seems to be just grounds for this belief. In the English asylums, however, the mortality is much lower, and is but little, if any, greater than that amongst the general population. This can be verified by examination of the Registrar-General's Report, from which it appears that phthisis accounts approximately for one in twelve of all deaths. These figures probably represent less than the true proportion of deaths from phthisis, since post-mortem examinations in the majority of cases are not made, and thus, doubtless, many cases of pulmonary tuberculosis are overlooked.

The true proportion of the circle to the square, circumscribed expressed in rational numbers, by Gothofredus Gulielmus Leebnitius, Englished from the Leipsick Acta Eruditorum Number II.

Geometricians, have always indeavoured to find out the Proportions between straight and Curve Lines.