BMP/SMAD signaling is a crucial regulator of intestinal differentiation 1 – 4 . However, the molecular underpinnings of the BMP pathway in this context are unknown. Here, we characterize the mechanism by which BMP/SMAD signaling drives enterocyte differentiation. We establish that the transcription factor HNF4A acts redundantly with an intestine-restricted HNF4 paralog, HNF4G, to activate enhancer chromatin and upregulate the majority of transcripts enriched in the differentiated epithelium; cells fail to differentiate upon double knockout of both HNF4 paralogs. Furthermore, we show that SMAD4 and HNF4 function via a reinforcing feed-forward loop, activating each other’s expression and co-binding to regulatory elements of differentiation genes. This feed-forward regulatory module promotes and stabilizes enterocyte cell identity; disruption of the HNF4-SMAD4 module results in loss of enterocyte fate in favor of progenitor and secretory cell lineages. This intersection of signaling and transcriptional control provides a framework to understand regenerative tissue homeostasis, particularly in tissues with inherent cellular plasticity 5 .

Summary Oncogenic mutations in BRAF are believed to initiate serrated colorectal cancers, however the mechanisms of BRAF-driven colon cancer are unclear. We find that oncogenic BRAF paradoxically suppresses stem cell renewal and instead promotes differentiation. Correspondingly, tumor formation is inefficient in BRAF-driven mouse models of colon cancer. By reducing levels of differentiation via genetic manipulation of either of two distinct differentiation-promoting factors (Smad4 or Cdx2), stem cell activity is restored in BRAFV600E intestines, and the oncogenic capacity of BRAFV600E is amplified. In human patients, we observe that reduced levels of differentiation in normal tissue is associated with increased susceptibility to serrated colon tumors. Together, these findings help resolve the conditions necessary for BRAF-driven colon cancer initiation. Additionally, our results predict that genetic and/or environmental factors which reduce tissue differentiation will increase susceptibility to serrated colon cancer. These findings offer an opportunity to identify susceptible individuals by assessing their tissue-differentiation status.

Model organisms are essential experimental platforms for discovering gene functions, defining protein and genetic networks, uncovering functional consequences of human genome variation, and for modeling human disease. For decades, researchers who use model organisms have relied on Model Organism Databases (MODs) and the Gene Ontology Consortium (GOC) for expertly curated annotations, and for access to integrated genomic and biological information obtained from the scientific literature and public data archives. Through the development and enforcement of data and semantic standards, these genome resources provide rapid access to the collected knowledge of model organisms in human readable and computation-ready formats that would otherwise require countless hours for individual researchers to assemble on their own. Since their inception, the MODs for the predominant biomedical model organisms [Mus sp. (laboratory mouse), Saccharomyces cerevisiae, Drosophila melanogaster, Caenorhabditis elegans, Danio rerio, and Rattus norvegicus] along with the GOC have operated as a network of independent, highly collaborative genome resources. In 2016, these six MODs and the GOC joined forces as the Alliance of Genome Resources (the Alliance). By implementing shared programmatic access methods and data-specific web pages with a unified “look and feel,” the Alliance is tackling barriers that have limited the ability of researchers to easily compare common data types and annotations across model organisms. To adapt to the rapidly changing landscape for evaluating and funding core data resources, the Alliance is building a modern, extensible, and operationally efficient “knowledge commons” for model organisms using shared, modular infrastructure.

AI is a broad concept, grouping initiatives that use a computer to perform tasks that would usually require a human to complete. AI methods are well suited to predict clinical outcomes. In practice, AI methods can be thought of as functions that learn the outcomes accompanying standardized input data to produce accurate outcome predictions when trialed with new data. Current methods for cleaning, creating, accessing, extracting, augmenting, and representing data for training AI clinical prediction models are well defined. The use of AI to predict clinical outcomes is a dynamic and rapidly evolving arena, with new methods and applications emerging. Extraction or accession of electronic health care records and combining these with patient genetic data is an area of present attention, with tremendous potential for future growth. Machine learning approaches, including decision tree methods of Random Forest and XGBoost, and deep learning techniques including deep multi-layer and recurrent neural networks, afford unique capabilities to accurately create predictions from high dimensional, multimodal data. Furthermore, AI methods are increasing our ability to accurately predict clinical outcomes that previously were difficult to model, including time-dependent and multi-class outcomes. Barriers to robust AI-based clinical outcome model deployment include changing AI product development interfaces, the specificity of regulation requirements, and limitations in ensuring model interpretability, generalizability, and adaptability over time.

No abstract