Paneth cells are the primary source of C-type lysozyme, a b-1,4-N-acetylmuramoylhydrolase that enzymatically processes bacterial cell walls. Paneth cells are normally present in human cecum and ascending colon, but are rarely found in descending colon and rectum; Paneth cell metaplasia in this region and aberrant lysozyme production are hallmarks of inflammatory bowel disease (IBD) pathology. Here, we examined the impact of aberrant lysozyme production in colonic inflammation. Targeted disruption of Paneth cell lysozyme (Lyz1) protected mice from experimental colitis. Lyz1-deficiency diminished intestinal immune responses to bacterial molecular patterns and resulted in the expansion of lysozyme-sensitive mucolytic bacteria, including Ruminococcus gnavus, a Crohn's disease-associated pathobiont. Ectopic lysozyme production in colonic epithelium suppressed lysozyme-sensitive bacteria and exacerbated colitis. Transfer of R. gnavus into Lyz1 À/À hosts elicited a type 2 immune response, causing epithelial reprograming and enhanced anti-colitogenic capacity. In contrast, in lysozyme-intact hosts, processed R. gnavus drove pro-inflammatory responses. Thus, Paneth cell lysozyme balances intestinal anti-and pro-inflammatory responses, with implications for IBD.

We propose an efficient neural framework for sentence-level discourse analysis in accordance with Rhetorical Structure Theory (RST). Our framework comprises a discourse segmenter to identify the elementary discourse units (EDU) in a text, and a discourse parser that constructs a discourse tree in a top-down fashion. Both the segmenter and the parser are based on Pointer Networks and operate in linear time. Our segmenter yields an F 1 score of 95.4, and our parser achieves an F 1 score of 81.7 on the aggregated labeled (relation) metric, surpassing previous approaches by a good margin and approaching human agreement on both tasks (98.3 and 83.0 F 1 ).

Clustering is a critical step in single cell-based studies. Most existing methods support unsupervised clustering without the a priori exploitation of any domain knowledge. When confronted by the high dimensionality and pervasive dropout events of scRNA-Seq data, purely unsupervised clustering methods may not produce biologically interpretable clusters, which complicates cell type assignment. In such cases, the only recourse is for the user to manually and repeatedly tweak clustering parameters until acceptable clusters are found. Consequently, the path to obtaining biologically meaningful clusters can be ad hoc and laborious. Here we report a principled clustering method named scDCC, that integrates domain knowledge into the clustering step. Experiments on various scRNA-seq datasets from thousands to tens of thousands of cells show that scDCC can significantly improve clustering performance, facilitating the interpretability of clusters and downstream analyses, such as cell type assignment.

Several immunotherapy clinical trials in recurrent glioblastoma have reported long-term survival benefits in 10–20% of patients. Here we perform genomic analysis of tumor tissue from recurrent WHO grade IV glioblastoma patients acquired prior to immunotherapy intervention. We report that very low tumor mutation burden is associated with longer survival after recombinant polio virotherapy or after immune checkpoint blockade in recurrent glioblastoma patients. A relationship between tumor mutation burden and survival is not observed in cohorts of immunotherapy naïve newly diagnosed or recurrent glioblastoma patients. Transcriptomic analyses reveal an inverse relationship between tumor mutation burden and enrichment of inflammatory gene signatures in cohorts of recurrent, but not newly diagnosed glioblastoma tumors, implying that a relationship between tumor mutation burden and tumor-intrinsic inflammation evolves upon recurrence.

Transition-based top-down parsing with pointer networks has achieved state-of-the-art results in multiple parsing tasks, while having a linear time complexity.However, the decoder of these parsers has a sequential structure, which does not yield the most appropriate inductive bias for deriving tree structures. In this paper, we propose hierarchical pointer network parsers, and apply them to dependency and sentence-level discourse parsing tasks. Our results on standard benchmark datasets demonstrate the effectiveness of our approach, outperforming existing methods and setting a new state-of-the-art.