In the early online edition of Genome Research this week, the University of Manchester's Andrew Sharrocks is lead author on a paper that used ChIP-chip to find all the target genes of the ETS-domain transcription factor ELK1. His team found that there are two modes by which ELK1 targets genes, and that one of the most "prominent groups of genes forming the ELK1 target network" are those that are part of the basal transcriptional machinery, the spliceosome, and the ribosome.
Michael Rehli at University Hospital Regensburg led work that shows that allele-specific DNA methylation in mouse strains is mainly determined by cis-acting sequences. Using a combination of "methyl-CpG immunoprecipitation and locus-wide tiling arrays," the scientists found several hundred differentially methylated regions between cells from two inbred mouse strains. Importantly, they write, "in F1 hybrids, the majority of strain-specific methylation patterns in somatic cells were maintained on the parental allele, regardless of their status in male germ line."
Over at the Wellcome Trust Sanger Institute, Heinrich Magnus Manske and Dominic Kwiatkowski have tackled visualizing high-throughput sequencing data with a new browser-based viewer called LookSeq, which is freely available. The AJAX-based Web viewer is for browsing large datasets of aligned reads, and it can render paired-end sequence reads in a way that is "more revealing about potential insertions and deletions than conventional methods," they say.
Also new in tools, scientists at the University of California, Berkeley, and UC Davis introduce a new model-based base-calling algorithm for Illumina called BayesCall. "This new approach significantly improves the accuracy over Illumina's basecaller Bustard, particularly in the later cycles of a sequencing run," they say. Testing it out on a standard viral sample, BayesCall improved on Bustard's average per-base error rate by about 52 percent.