Maximizing the Power of Principal-Component Analysis of Correlated Phenotypes in Genome-wide Association Studies

Many human traits are highly correlated. This correlation can be leveraged to improve the power of genetic association tests to identify markers associated with one or more of the traits. Principal component analysis (PCA) is a useful tool that has been widely used for the multivariate analysis of correlated variables. PCA is usually applied as a dimension reduction method: the few top principal components (PCs) explaining most of total trait variance are tested for association with a predictor of interest, and the remaining components are not analyzed. In this study we review the theoretical basis of PCA and describe the behavior of PCA when testing for association between a SNP and correlated traits. We then use simulation to compare the power of various PCA-based strategies when analyzing up to 100 correlated traits. We show that contrary to widespread practice, testing only the top PCs often has low power, whereas combining signal across all PCs can have greater power. This power gain is primarily due to increased power to detect genetic variants with opposite effects on positively correlated traits and variants that are exclusively associated with a single trait. Relative to other methods, the combined-PC approach has close to optimal power in all scenarios considered while offering more flexibility and more robustness to potential confounders. Finally, we apply the proposed PCA strategy to the genome-wide association study of five correlated coagulation traits where we identify two candidate SNPs that were not found by the standard approach.     

Recombination in the Human Pseudoautosomal Region PAR1

The pseudoautosomal region (PAR) is a short region of homology between the mammalian X and Y chromosomes, which has undergone rapid evolution. A crossover in the PAR is essential for the proper disjunction of X and Y chromosomes in male meiosis, and PAR deletion results in male sterility. This leads the human PAR with the obligatory crossover, PAR1, to having an exceptionally high male crossover rate, which is 17-fold higher than the genome-wide average. However, the mechanism by which this obligatory crossover occurs remains unknown, as does the fine-scale positioning of crossovers across this region. Recent research in mice has suggested that crossovers in PAR may be mediated independently of the protein PRDM9, which localises virtually all crossovers in the autosomes. To investigate recombination in this region, we construct the most fine-scale genetic map containing directly observed crossovers to date using African-American pedigrees. We leverage recombination rates inferred from the breakdown of linkage disequilibrium in human populations and investigate the signatures of DNA evolution due to recombination. Further, we identify direct PRDM9 binding sites using ChIP-seq in human cells. Using these independent lines of evidence, we show that, in contrast with mouse, PRDM9 does localise peaks of recombination in the human PAR1. We find that recombination is a far more rapid and intense driver of sequence evolution in PAR1 than it is on the autosomes. We also show that PAR1 hotspot activities differ significantly among human populations. Finally, we find evidence that PAR1 hotspot positions have changed between human and chimpanzee, with no evidence of sharing among the hottest hotspots. We anticipate that the genetic maps built and validated in this work will aid research on this vital and fascinating region of the genome.     

Targeting conserved water molecules: Design of 4-aryl-5-cyanopyrrolo[2,3-d]pyrimidine Hsp90 inhibitors using fragment-based screening and structure-based optimization

Inhibitors of the Hsp90 molecular chaperone are showing promise as anti-cancer agents. Here we describe a series of 4-aryl-5-cyanopyrrolo[2,3-d]pyrimidine ATP competitive Hsp90 inhibitors that were identified following structure-driven optimization of purine hits revealed by NMR based screening of a proprietary fragment library. Ligand-Hsp90 X-ray structures combined with molecular modeling led to the rational displacement of a conserved water molecule leading to enhanced affinity for Hsp90 as measured by fluorescence polarization, isothermal titration calorimetry and surface plasmon resonance assays. This displacement was achieved with a nitrile group, presenting an example of efficient gain in binding affinity with minimal increase in molecular weight. Some compounds in this chemical series inhibit the proliferation of human cancer cell lines in vitro and cause depletion of oncogenic Hsp90 client proteins and concomitant elevation of the co-chaperone Hsp70. In addition, one compound was demonstrated to be orally bioavailable in the mouse. This work demonstrates the power of structure-based design for the rapid evolution of potent Hsp90 inhibitors and the importance of considering conserved water molecules in drug design.     

An endopeptidase associated with bovine neurohypophysis secretory granules cleaves pro-ocytocin/neurophysin peptide at paired basic residues

The octacosapeptide sequence [Tyr18] pro-ocytocin/neurophysin (1-18)NH2 [pro-OT/Np(1-18)NH2] was synthesized and used as substrate to detect endoprotease(s) possibly involved in the processing of this precursor in bovine hypothalamo-neurohypophyseal tract. An endopeptidase (58 Kda) was detected in Lysates made from highly purified neurosecretory granules. This protease which cleaves the peptide bond on the carboxyl side of the Lys-Arg doublet, and no single basic residue, generates both OT-Gly10-Lys11-Arg12+Ala13-Val-Leu-Asp-Leu-Tyr18 (NH2) from the octacosapeptide substrate. In addition, a carboxypeptidase B-like activity converting OT-Gly10-Lys11-Arg12 into OT-Gly10 was detected in the same granule Lysates. It is hypothesized that a combination of these endoprotease and carboxypeptidase B-like activities together with the amidating enzyme of secretory granules might participate in the cleavage and processing of pro-OT/Np in vivo.     

Consequences of adaptive foraging in diverse communities

1.  Selective pressures acting on foraging activities constrain the strength of interaction, hence the stability and energetic availability in food webs.
2.  Because such selective pressures are usually measured at the individual level and because most experimental and theoretical works focus on simple settings, linking adaptive foraging with community scale patterns is still a far stretch.
3.  Some recent models incorporate foraging adaptation in diverse communities. The models vary in the way they incorporate adaptation, via evolutionary or behavioural changes, and define individual fitness in various ways.
4.  In spite of these differences, some general results linking adaptation to community structure and functioning emerge. In the present article, I introduce these different models and highlight their common results.
5.  Adaptive foraging provides stability to large food web models and predicts successfully interaction patterns within food webs as well as other topological features such as food chain length.
6.  The relationships between adaptive foraging and other structuring factors particularly depend on how well connected the local community is with surrounding communities (metacommunity aspect).