>> Settling an oocyte actin conundrum
Actin is actively cleared from the nuclei of most eukaryotic cells with the exception of the giant nuclei found in amphibian oocytes. This discrepancy has led to speculations about exotic actin conformations serving equally exotic functions in specialized nuclei. Dirk Görlich of the Center for Molecular Biology at Heidelberg University and his group recently reported that the mechanism for having actin in these oocyte nuclei is actually quite simple.
"I've found the idea of whether there's actin in the nucleus to be interesting...
A program to improve experimental design
Statistical power analyses help researchers to decide on the most effective number of subjects to test prior to study. But such decisions in microarray experiments are often made based on factors like cost and feasibility. Eric Hoffman and colleagues at the Children's National Medical Center in Washington, DC, developed a free Web tool for performing power analysis using pre-existing Affymetrix microarray data.
The Hierarchical Clustering Explorer 3.5 (www.cnmcresearch.org/bioinformatics), which probes the effects of varying significance, error rate, sample size, and effect size, deals with what Charles Auffray, Faculty of 1000 member at the CNRS calls, "an absolutely essential issue that has been disregarded for too long."
"This paper is providing a solution. It's not perfect, of course, but it makes it possible for a global and systematic assessment of the power of a microarray study.
"One of the problems with the Affymetrix chips is that because they use a set of short probes for any given gene they collect signals that differ a lot from each other. They then combine them to get what they call a present call or absent call. This program provides a means to assess which are the set of probes on the chip for which you can have trust. Beyond just implementing the statistical power analyses, this is something that is very important."1. J. Seo et al., "An interactive power analysis tool for microarray hypothesis testing and generation," Bioinformatics, 22:808-14, April 1, 2006.