From promoter sequence to expression

摘要: We present a probabilistic framework that models the process by which transcriptional binding explains mRNA expression of different genes. Our joint model unifies two key components this process: prediction gene regulation events from sequence motifs in gene's promoter region, and combinations settings. approach has several advantages. By learning are directly predictive data, it can improve identification site patterns. It is also able to identify combinatorial via interactions transcription factors. Finally, general allows us integrate additional data sources, including recent localization assays. demonstrate our on cell cycle Spellman et al., combined with information Simon al. show learned predicts sequence, identifies coherent co-regulated groups significant factor motifs. provides valuable biological insight into domain these "modules" effects govern their behavior.