A powerful method for pleiotropic analysis under composite null hypothesis identifies novel shared loci between Type 2 Diabetes and Prostate Cancer

There is increasing evidence that pleiotropy, the association of multiple traits with the same genetic variants/loci, is a very common phenomenon. Cross-phenotype association tests are often used to jointly analyze multiple traits from a genome-wide association study (GWAS). The underlying methods, however, are often designed to test the global null hypothesis that there is no association of a genetic variant with any of the traits, the rejection of which does not implicate pleiotropy. In this article, we propose a new statistical approach, PLACO, for specifically detecting pleiotropic loci between two traits by considering an underlying composite null hypothesis that a variant is associated with none or only one of the traits. We propose testing the null hypothesis based on the product of the Z-statistics of the genetic variants across two studies and derive a null distribution of the test statistic in the form of a mixture distribution that allows for fractions of variants to be associated with none or only one of the traits. We borrow approaches from the statistical literature on mediation analysis that allow asymptotic approximation of the null distribution avoiding estimation of nuisance parameters related to mixture proportions and variance components. Simulation studies demonstrate that the proposed method can maintain type I error and can achieve major power gain over alternative simpler methods that are typically used for testing pleiotropy. PLACO allows correlation in summary statistics between studies that may arise due to sharing of controls between disease traits. Application of PLACO to publicly available summary data from two large case-control GWAS of Type 2 Diabetes and of Prostate Cancer implicated a number of novel shared genetic regions: 3q23 (ZBTB38), 6q25.3 (RGS17), 9p22.1 (HAUS6), 9p13.3 (UBAP2), 11p11.2 (RAPSN), 14q12 (AKAP6), 15q15 (KNL1) and 18q23 (ZNF236).     

Outlier Removal and the Relation with Reporting Errors and Quality of Psychological Research

Background

The removal of outliers to acquire a significant result is a questionable research practice that appears to be commonly used in psychology. In this study, we investigated whether the removal of outliers in psychology papers is related to weaker evidence (against the null hypothesis of no effect), a higher prevalence of reporting errors, and smaller sample sizes in these papers compared to papers in the same journals that did not report the exclusion of outliers from the analyses.

Methods and Findings

We retrieved a total of 2667 statistical results of null hypothesis significance tests from 153 articles in main psychology journals, and compared results from articles in which outliers were removed (N = 92) with results from articles that reported no exclusion of outliers (N = 61). We preregistered our hypotheses and methods and analyzed the data at the level of articles. Results show no significant difference between the two types of articles in median p value, sample sizes, or prevalence of all reporting errors, large reporting errors, and reporting errors that concerned the statistical significance. However, we did find a discrepancy between the reported degrees of freedom of t tests and the reported sample size in 41% of articles that did not report removal of any data values. This suggests common failure to report data exclusions (or missingness) in psychological articles.

Conclusions

We failed to find that the removal of outliers from the analysis in psychological articles was related to weaker evidence (against the null hypothesis of no effect), sample size, or the prevalence of errors. However, our control sample might be contaminated due to nondisclosure of excluded values in articles that did not report exclusion of outliers. Results therefore highlight the importance of more transparent reporting of statistical analyses.     

Biorhythms and tests of oral skills

The null hypothesis that there is no difference in an individuaľs level of skill on critical and non-critical days of the biorhythm cycle is tested over a 2-week period using oral stereognosis and a dysarthria assessment. No statistical evidence could be found to refute the null hypothesis proposed and it is concluded that there is no relationship between oral skills and stages of the biorhythm cycle.     

A Bayesian Perspective on the Reproducibility Project: Psychology

We revisit the results of the recent Reproducibility Project: Psychology by the Open Science Collaboration. We compute Bayes factors—a quantity that can be used to express comparative evidence for an hypothesis but also for the null hypothesis—for a large subset (N = 72) of the original papers and their corresponding replication attempts. In our computation, we take into account the likely scenario that publication bias had distorted the originally published results. Overall, 75% of studies gave qualitatively similar results in terms of the amount of evidence provided. However, the evidence was often weak (i.e., Bayes factor < 10). The majority of the studies (64%) did not provide strong evidence for either the null or the alternative hypothesis in either the original or the replication, and no replication attempts provided strong evidence in favor of the null. In all cases where the original paper provided strong evidence but the replication did not (15%), the sample size in the replication was smaller than the original. Where the replication provided strong evidence but the original did not (10%), the replication sample size was larger. We conclude that the apparent failure of the Reproducibility Project to replicate many target effects can be adequately explained by overestimation of effect sizes (or overestimation of evidence against the null hypothesis) due to small sample sizes and publication bias in the psychological literature. We further conclude that traditional sample sizes are insufficient and that a more widespread adoption of Bayesian methods is desirable.     

Organization and Evolution of the Mammalian Genome: I. Polymorphism of H-2 Linked Loci

To test the hypothesis that the H-2 polymorphism is adaptive, the degree of polymorphism of loci linked to the H-2 complex on chromosome 17 of the house mouse was compared to the degree of polymorphism of loci located on other chromosomes. Published theoretical analyses show that polymorphisms subject to natural selection usually reduce the polymorphism of linked neutral loci. The first test of the hypothesis was based on data obtained from a survey of the polymorphism of 12 isozyme-encoding loci in wild house mice from Europe, North Africa and South America. Results of this test showed that, on the average, H-2-linked loci were as polymorphic as loci located on other chromosomes. In fact, the data suggested that H-2 linked loci might be more polymorphic than other loci. To test this hypothesis more rigorously, data for the 12 isozyme-encoding loci were augmented with data from published surveys of the polymorphisms of 59 loci in house mice from Europe and North America. Results of these tests showed that polymorphic loci linked to the H-2 complex tended to be more, rather than less, polymorphic than loci located on other chromosomes. The cluster of highly polymorphic loci seems to be related to linkage of these loci to the highly polymorphic H-2 complex, but the way in which the influence is exerted could not be readily explained.     

Biochemical evidence for a separate,MHC-linked locus encoding H-2.28 antigens

In comparing the tryptic peptide maps of the H-2L and H-2D glycoprotein antigens isolated from NP-40 lysates of RADA1 (H-2 ^a ) leukemic cells, no more than 37% of the observed arginine-containing tryptic peptides are found to be homologous. Thus, the primary amino-acid sequences of these two antigens are probably less than 90% homologous. This constitutes the strongest evidence to date that the MHC-linkedH-2L region encodes H-2L antigens separately from theH-2D region, even though H-2L antigens bear D-end-associated antigenic determinants of the H-2.28 family. The anti-H-2.28 alloantiserum (k×r anti h2) used to precipitate H-2L antigens in this investigation was the NIH contract antiserum D28b. As the tryptic peptide maps also surprisingly revealed, D28b precipitates H-2D antigens as well and, thus, anti-H-2.4 immunoadsorbants were employed to isolate H-2L free of H-2D antigens. In light of the dual specificity of D28b, its reactivity with BALB/c-H-2 ^dm2 mutant cells was re-examined. Even though mutant lymphocytes, which lack H-2L but not H-2D antigens, are not cytotoxically lysed by D28b (as are parental H-2^d cells), D28b appears to precipitate H-2D antigens from NP-40 extracts of mutant splenocytes.     

The association of H-2 haplotype with implantation,survival, and growth of murine embryos

Using three congenic strains, C57BL/10Sn (H-2 ^b), B10.A/SnSg (H-2 ^a), and B10.D2/nSn (H-2 ^d), we sought to investigate the possible association of H-2 haplotype with the number of implants, fetal survival, and fetal weight, as well as to analyze the possible effects of hybrid vigor and maternal-fetal histoincompatibility in primigravidae mice. The results of this study indicate a significant association between genes at or near the H-2 complex and both fetal loss and fetal weight, but not the number of implants. Haplotype variation accounted for 14 percent of the variation in fetal loss and 20 percent of the variation in fetal weight. With the exception of fetal loss, there was no evidence of a maternal effect. There was also no clear evidence of hybrid vigor or histoincompatibility effects for any of the three variables studied. In summary, the data suggests that particular allelic variants at or near the H-2 complex confer some selective advantage as measured by differential fetal survival and fetal growth.     

Do yeasts and Drosophila interact just by chance?

The fruit fly Drosophila melanogaster and the baker's yeast Saccharomyces cerevisiae are classic research model organisms that are also associated in nature, at least around vineyards. Sharing the same ephemeral fruit niche, winged Drosophila feed on immotile yeasts. That a yeast diet is essential for larval development, and that saprophagous fruit flies are attracted to a suite of yeast volatiles, has been well established over the last century. Recently, research has focussed on the potential mutual benefit of this interaction hypothesising yeasts also benefit via dispersal from ephemeral fruits. It now appears that the concept of a co-evolved mutualism between yeasts and Drosophila has permeated the literature. However, until robust evidence regarding the evolution and maintenance of this yeast-fly association has been provided, we suggest there is no compelling evidence to reject the more simplistic null hypothesis that these interactions are due to exaptation, and not a mutualism driven by natural selection.     

P > .05: The incorrect interpretation of “not significant” results is a significant problem

Statistically nonsignificant (p > .05) results from a null hypothesis significance test (NHST) are often mistakenly interpreted as evidence that the null hypothesis is true—that there is “no effect” or “no difference.” However, many of these results occur because the study had low statistical power to detect an effect. Power below 50% is common, in which case a result of no statistical significance is more likely to be incorrect than correct. The inference of “no effect” is not valid even if power is high. NHST assumes that the null hypothesis is true; p is the probability of the data under the assumption that there is no effect. A statistical test cannot confirm what it assumes. These incorrect statistical inferences could be eliminated if decisions based on p values were replaced by a biological evaluation of effect sizes and their confidence intervals. For a single study, the observed effect size is the best estimate of the population effect size, regardless of the p value. Unlike p values, confidence intervals provide information about the precision of the observed effect. In the biomedical and pharmacology literature, methods have been developed to evaluate whether effects are “equivalent,” rather than zero, as tested with NHST. These methods could be used by biological anthropologists to evaluate the presence or absence of meaningful biological effects. Most of what appears to be known about no difference or no effect between sexes, between populations, between treatments, and other circumstances in the biological anthropology literature is based on invalid statistical inference.     

Selective Decline of Synaptic Protein Levels in the Frontal Cortex of Female Mice Deficient in the Extracellular Metalloproteinase ADAMTS1

The chondroitin sulfate-bearing proteoglycans, also known as lecticans, are a major component of the extracellular matrix (ECM) in the central nervous system and regulate neural plasticity. Growing evidence indicates that endogenous, extracellular metalloproteinases that cleave lecticans mediate neural plasticity by altering the structure of ECM aggregates. The bulk of this in vivo data examined the matrix metalloproteinases, but another metalloproteinase family that cleaves lecticans, a disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS), modulates structural plasticity in vitro, although few in vivo studies have tested this concept. Thus, the purpose of this study was to examine the neurological phenotype of a mouse deficient in ADAMTS1. Adamts1 mRNA was absent in the ADAMTS1 null mouse frontal cortex, but there was no change in the abundance or proteolytic processing of the prominent lecticans brevican and versican V2. However, there was a marked increase in the perinatal lectican neurocan in juvenile ADAMTS1 null female frontal cortex. More prominently, there were declines in synaptic protein levels in the ADAMTS1 null female, but not male, frontal cortex beginning at postnatal day 28. These synaptic marker declines did not affect learning or memory in the adult female ADAMTS1 null mice when tested with the radial-arm water maze. These results indicate that in vivo Adamts1 knockout leads to sexual dimorphism in frontal cortex synaptic protein levels. Since changes in lectican abundance and proteolytic processing did not accompany the synaptic protein declines, ADAMTS1 may play a nonproteolytic role in regulating neural plasticity.     

Relationships between private and public H-2 specificities on the cell surface

Differential redistribution was used to investigate relationships between private specificity H-2.4 and public specificity H-2.28, in the product of aD region allele of theH-2 complex. Monospecific anti-H-2 antisera and fluorochrome conjugated antimouse Ig antibodies were used to induce redistribution of H-2 antigens on the surface of peripheral T lymphocytes fromH-2 ^a andH-2 ^d mice. Results showed that redistribution of specificity H-2.4 into patches and caps did not induce concomittant redistribution of specificity H-2.28, which remain diffusely scattered on the cell surface outside the caps of H-2.4. Redistribution of H-2.28 induced redistribution of H-2.4, which was no longer detectable outside the caps of H-2.28. These data indicate that (a) at least some of the H-2.28 sites are expressed on polypeptide chains independent from those carrying H-2.4 and (b) other H-2.28 sites may be linked to molecules carrying H-2.4. Since, onH-2 ^a cells, both specificities are products of the D region of theH-2 gene complex, our results suggest that there are at least two genes in theD region.     

Applications of the Wei-Lachin Multivariate One-Sided Test for Multiple Outcomes on Possibly Different Scales

Many studies aim to assess whether a therapy has a beneficial effect on multiple outcomes simultaneously relative to a control. Often the joint null hypothesis of no difference for the set of outcomes is tested using separate tests with a correction for multiple tests, or using a multivariate T ²-like MANOVA or global test. However, a more powerful test in this case is a multivariate one-sided or one-directional test directed at detecting a simultaneous beneficial treatment effect on each outcome, though not necessarily of the same magnitude. The Wei-Lachin test is a simple 1 df test obtained from a simple sum of the component statistics that was originally described in the context of a multivariate rank analysis. Under mild conditions this test provides a maximin efficient test of the null hypothesis of no difference between treatment groups for all outcomes versus the alternative hypothesis that the experimental treatment is better than control for some or all of the component outcomes, and not worse for any. Herein applications are described to a simultaneous test for multiple differences in means, proportions or life-times, and combinations thereof, all on potentially different scales. The evaluation of sample size and power for such analyses is also described. For a test of means of two outcomes with a common unit variance and correlation 0.5, the sample size needed to provide 90% power for two separate one-sided tests at the 0.025 level is 64% greater than that needed for the single Wei-Lachin multivariate one-directional test at the 0.05 level. Thus, a Wei-Lachin test with these operating characteristics is 39% more efficient than two separate tests. Likewise, compared to a T ²-like omnibus test on 2 df, the Wei-Lachin test is 32% more efficient. An example is provided in which the Wei-Lachin test of multiple components has superior power to a test of a composite outcome.     

Pattern analysis in bryozoan and spirorbid communities. II. Distance sampling methods

The spatial pattern of settlements of the bryozoan Alcyonidium polyoum (Hassall) and of Spirorbis spirorbis (L.) on plastic (Perspex) disks has been examined using nearest-neighbour analysis. Cumulative frequency distributions of ω = r², where r is the distance from each individual to its nearest neighbour, have been set up and compared with those expected assuming random settlement on each disk. Deviation above the expected line indicates aggregation and deviation below it indicates regularity or spacing apart. At low densities (< 7.4 cm^?2) Spirorbis settlements were demonstrably aggregated and at high densities (> 18.2 cm^?2) they displayed spacing apart. Alcyonidium ancestrulae showed aggregation at all densities (1.4–45.6 cm^?2) under the null hypothesis that when a larva encounters an ancestrula it moves away and settles randomly elsewhere. Truncating the cumulative distributions to consider only the settlement within aggregations still showed no evidence of spacing apart. Under a different null hypothesis, that a larva encountering an ancestrula settles contiguously to it, spacing apart was demonstrated only at unnaturally high densities (> 13 cm^?2). It is concluded, on the basis of observed larval exploratory behaviour and naturally occurring settlement densities, that settling Alcyonidium larvae do not space apart. The biological significance of this finding is discussed.     

PSSA-2, a Membrane-Spanning Phosphoprotein of Trypanosoma brucei,Is Required for Efficient Maturation of Infection

The coat of Trypanosoma brucei consists mainly of glycosylphosphatidylinositol-anchored proteins that are present in several million copies and are characteristic of defined stages of the life cycle. While these major components of the coats of bloodstream forms and procyclic (insect midgut) forms are well characterised, very little is known about less abundant stage-regulated surface proteins and their roles in infection and transmission. By creating epitope-tagged versions of procyclic-specific surface antigen 2 (PSSA-2) we demonstrated that it is a membrane-spanning protein that is expressed by several different life cycle stages in tsetse flies, but not by parasites in the mammalian bloodstream. In common with other membrane-spanning proteins in T. brucei, PSSA-2 requires its cytoplasmic domain in order to exit the endoplasmic reticulum. Correct localisation of PSSA-2 requires phosphorylation of a cytoplasmic threonine residue (T₃₀₅), a modification that depends on the presence of TbMAPK4. Mutation of T₃₀₅ to alanine (T₃₀₅A) has no effect on the localisation of the protein in cells that express wild type PSSA-2. In contrast, this protein is largely intracellular when expressed in a null mutant background. A variant with a T₃₀₅D mutation gives strong surface expression in both the wild type and null mutant, but slows growth of the cells, suggesting that it may function as a dominant negative mutant. The PSSA-2 null mutant exhibits no perceptible phenotype in culture and is fully competent at establishing midgut infections in tsetse, but is defective in colonising the salivary glands and the production of infectious metacyclic forms. Given the protein''s structure and the effects of mutation of T₃₀₅ on proliferation and localisation, we postulate that PSSA-2 might sense and transmit signals that contribute to the parasite''s decision to divide, differentiate or migrate.     

Inactivation of mouse transmembrane prolyl 4-hydroxylase increases blood brain barrier permeability and ischemia-induced cerebral neuroinflammation

Hypoxia-inducible factor prolyl 4-hydroxylases (HIF-P4Hs) regulate the hypoxic induction of >300 genes required for survival and adaptation under oxygen deprivation. Inhibition of HIF-P4H-2 has been shown to be protective in focal cerebral ischemia rodent models, while that of HIF-P4H-1 has no effects and inactivation of HIF-P4H-3 has adverse effects. A transmembrane prolyl 4-hydroxylase (P4H-TM) is highly expressed in the brain and contributes to the regulation of HIF, but the outcome of its inhibition on stroke is yet unknown. To study this, we subjected WT and P4htm^−/− mice to permanent middle cerebral artery occlusion (pMCAO). Lack of P4H-TM had no effect on lesion size following pMCAO, but increased inflammatory microgliosis and neutrophil infiltration was observed in the P4htm^−/− cortex. Furthermore, both the permeability of blood brain barrier and ultrastructure of cerebral tight junctions were compromised in P4htm^−/− mice. At the molecular level, P4H-TM deficiency led to increased expression of proinflammatory genes and robust activation of protein kinases in the cortex, while expression of tight junction proteins and the neuroprotective growth factors erythropoietin and vascular endothelial growth factor was reduced. Our data provide the first evidence that P4H-TM inactivation has no protective effect on infarct size and increases inflammatory microgliosis and neutrophil infiltration in the cortex at early stage after pMCAO. When considering HIF-P4H inhibitors as potential therapeutics in stroke, the current data support that isoenzyme-selective inhibitors that do not target P4H-TM or HIF-P4H-3 would be preferred.     

Immune response of mice to Thy-1.1 antigen: Studies on congenic lines

The primary response to Thy-1.1 antigen was measured by a plaque assay that detected cells producing antibodies lytic for AKR thymocytes (PFC). TheH-2 congenic mice (B10.K and B10.BR) carryingH-2 complexes of high responders (CBA and C57BR) on the low-responder background (B10) were found to produce significantly fewer PFC than the corresponding donor of theH-2 complex. On the other hand, C3H.B10 mice carrying theH-2 complex of a low responder on the high-responder background produced significantly more PFC than the donor of theH-2 complex. These findings were interpreted as evidence that alleles at previously described loci believed to be components of theI region of theH-2 complex and controlling immune response to Thy-1.1 are influenced by alleles at another locus. Studies of segregating populations of theH-2 congenic lines supplied evidence that this locus, tentatively calledIr-5, is in chromosome 17 (linkage group IX).     

Significant association between GSTT1 null genotype and risk of asthma during childhood in Caucasians

Asthma is a complex multifactorial disorder and its management requires a better understanding of its various pathogenesis and mechanisms. Previous studies assessing the association between glutathione S-transferase T1 (GSTT1) null genotype and asthma risk during childhood reported conflicting results. To get a more precise estimation of the association between GSTT1 null genotype and risk of asthma during childhood, we performed a meta-analysis of 16 studies with a total of 18,558 subjects. Subgroup analyses were performed by ethnicity. The pooled odds ratio (OR) with corresponding 95 % confidence interval (95 %CI) was used to assess the association. Overall, there was a significant association between GSTT1 null genotype and increased risk of children asthma (OR = 1.25, 95 % CI, 1.02–1.54; P = 0.032). Subgroup analyses showed GSTT1 null genotype was associated with increased risk of children asthma in Caucasians (OR = 1.46, 95 % CI, 1.04–2.03; P = 0.027), but not in Asians (OR = 1.03, 95 % CI, 0.55–1.94; P = 0.928) and Africans (OR = 1.33, 95 % CI, 0.92–1.91; P = 0.127). There was no evidence of publication bias in the subgroup analysis of Caucasians. In conclusion, there is a significant association between GSTT1 null genotype and risk of asthma during childhood in Caucasians. More well-designed epidemiological studies are needed to further assess this association in Asians and Africans.     

Tests for Genetic Differentiation

When testing for genetic differentiation the joint null hypothesis that there is no allele frequency difference at any locus is of interest. Common approaches to test this hypothesis are based on the summation of χ² statistics over loci and on the Bonferroni correction, respectively. Here, we also consider the Simes adjustment and a recently proposed truncated product method (TPM) to combine P‐values. The summation and the TPM (using a relatively large truncation point) are powerful when there are differences in many or all loci. The Simes adjustment, however, is powerful when there are differences regarding one or a few loci only. As a compromise between the different approaches we introduce a combination between the Simes adjustment and the TPM, i.e. the joint null hypothesis is rejected if at least one of the two methods, Simes and TPM, is significant at the α/2‐level. Simulation results indicate that this combination is a robust procedure with high power over the different types of alternatives.     

APOL1 Null Alleles from a Rural Village in India Do Not Correlate with Glomerulosclerosis

Background

Among African-Americans, genome wide association revealed a strong correlation between the G1 and G2 alleles of APOL1 (apolipoproteinL1, also called trypanolytic factor) and kidney diseases including focal and segmental glomerulosclerosis, HIV-associated nephropathy and hypertensive nephrosclerosis. In the prevailing hypothesis, heterozygous APOL1 G1 and G2 alleles increase resistance against Trypanosoma that cause African sleeping sickness, resulting in positive selection of these alleles, but when homozygous the G1 and G2 alleles predispose to glomerulosclerosis. While efforts are underway to screen patients for G1 and G2 alleles and to better understand “APOL1 glomerulopathy,” no data prove that these APOL1 sequence variants cause glomerulosclerosis. G1 and G2 correlate best with glomerulosclerosis as recessive alleles, which suggests a loss of function mutation for which proof of causality is commonly tested with homozygous null alleles. This test cannot be performed in rodents as the APOL gene cluster evolved only in primates. However, there is a homozygous APOL1 null human being who lives in a village in rural India. This individual and his family offer a unique opportunity to test causality between APOL1 null alleles and glomerulosclerosis.

Methods and Findings

We obtained clinical data, blood and urine from this APOL1 null patient and 50 related villagers. Based on measurements of blood pressure, BUN, creatinine, albuminuria, genotyping and immunoblotting, this APOL1 null individual does not have glomerulosclerosis, nor do his relatives who carry APOL1 null alleles.

Conclusions

This small study cannot provide definitive conclusions but the absence of glomerulosclerosis in this unique population is consistent with the possibility that African-American glomerulosclerosis is caused, not by loss of APOL1 function, but by other mechanisms including a subtle gain of function or by the “genetic hitchhiking” of deleterious mutations in a gene linked to APOL1 G1 and G2.