A DNA methylation reader complex that enhances gene transcription
DNA methylation generally functions as a repressive transcriptional signal, but it is also known to activate gene expression. In either case, the downstream factors remain largely unknown. By using comparative interactomics, we isolated proteins in Arabidopsis thaliana that associate with methylated DNA. Two SU(VAR)3-9 homologs, the transcriptional antisilencing factor SUVH1, and SUVH3, were among the methyl reader candidates. SUVH1 and SUVH3 bound methylated DNA in vitro, were associated with euchromatic methylation in vivo, and formed a complex with two DNAJ domain-containing homologs, DNAJ1 and DNAJ2. Ectopic recruitment of DNAJ1 enhanced gene transcription in plants, yeast, and mammals. Thus, the SUVH proteins bind to methylated DNA and recruit the DNAJ proteins to enhance proximal gene expression, thereby counteracting the repressive effects of transposon insertion near genes.
Early human dispersals within the Americas
Studies of the peopling of the Americas have focused on the timing and number of initial migrations. Less attention has been paid to the subsequent spread of people within the Americas. We sequenced 15 ancient human genomes spanning from Alaska to Patagonia; six are ≥10,000 years old (up to ~18x coverage).
LZTR1 is a regulator of RAS ubiquitination and signaling
In genetic screens aimed at understanding drug resistance mechanisms in chronic myeloid leukemia cells, inactivation of the cullin 3 adapter protein-encoding leucine zipper-like transcription regulator 1 (LZTR1) gene led to enhanced mitogen-activated protein kinase (MAPK) pathway activity and reduced sensitivity to tyrosine kinase inhibitors. Knockdown of the Drosophila LZTR1 ortholog CG3711 resulted in a Ras-dependent gain-of-function phenotype. Endogenous human LZTR1 associates with the main RAS isoforms.
A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play
The game of chess is the longest-studied domain in the history of artificial intelligence. The strongest programs are based on a combination of sophisticated search techniques, domain-specific adaptations, and handcrafted evaluation functions that have been refined by human experts over several decades. By contrast, the AlphaGo Zero program recently achieved superhuman performance in the game of Go by reinforcement learning from self-play. In this paper, we generalize this approach into a single AlphaZero algorithm that can achieve superhuman performance in many challenging games.