Cross-species cell-type assignment of single-cell RNA-seq by a heterogeneous graph neural network

Cross-species comparative analyses of single-cell RNA sequencing (scRNA-seq) data allow us to explore, at single-cell resolution, the origins of cellular diversity and the evolutionary mechanisms that shape cellular form and function. Here, we aimed to utilize a heterogeneous graph neural network to learn aligned and interpretable cell and gene embeddings for cross-species cell type assignment and gene module extraction (CAME) from scRNA-seq data. A systematic evaluation study on 649 pairs of cross-species datasets showed that CAME outperformed six benchmarking methods in terms of cell-type assignment and model robustness to insufficiency and inconsistency of sequencing depths. Comparative analyses of the major types of human and mouse brains by CAME revealed shared cell type-specific functions in homologous gene modules. Alignment of the trajectories of human and macaque spermatogenesis by CAME revealed conservative gene expression dynamics during spermatogenesis between humans and macaques. Owing to the utilization of non-one-to-one homologous gene mappings, CAME made a significant improvement on cell-type characterization cross zebrafish and other species. Overall, CAME can not only make an effective cross-species assignment of cell types on scRNA-seq data but also reveal evolutionary conservative and divergent features between species.

OTME-3. Dissection of the role of stromal microenvironment and tumor-TME crosstalk in pediatric brain cancer

Brain tumors are the deadliest malignancies that occur during childhood and strong efforts are required to develop innovative therapeutic strategies. The intrinsic capacity of malignant cells to organize, shape and exploit the surrounding environment where they develop (tumor microenvironment, TME), has not been fully elucidated for pediatric brain cancers yet. Here, we exploited a multi-omic approach to define the TME cell populations and their contributions in the most common pediatric brain tumor entities, such as medulloblastomas and ependymomas. Analysis of single-cell RNA sequencing data of human tumors resulted in the identification of heterogeneous populations of non-malignant cells present in the TME. In particular, re-clustering and marker-based cell type assignment strategies allowed to define a broad range of immune and stromal subclasses showing distinctive expression signatures reflecting variegated functional roles. By cross-matching the tumor data with normal brain expression atlases, we could further refine the annotation of the newly identified stromal functional subpopulations and define the “tumor-associated” marker signatures of genes exclusively enriched in stromal cells within the TME, linked to immune activation, cell adhesion and cytokine regulation pathways. Bulk transcriptomic data of human tumors and matching patient-derived xenografts (PDXs) showed that a group of secreted stromal factors acting as regulators of tumorigenic mechanisms, such as IGF2 and COL4A1, are lost after xenografting and replaced by the host murine microenvironment, suggesting that tumor cells are involved in paracrine and bivalent crosstalk with TME cells, impacting on tumor cell growth and progression. Finally, bulk deconvolution and cell-cell communication analysis were exploited to define, respectively, the stromal cell proportions and the key factors involved in the tumor-TME crosstalk; this latter can be considered as possible targets for tailored and more specific anti-tumor therapeutic strategies.

Kripke Semantics for Intersection Formulas

We propose a notion of the Kripke-style model for intersection logic. Using a game interpretation, we prove soundness and completeness of the proposed semantics. In other words, a formula is provable (a type is inhabited) if and only if it is forced in every model. As a by-product, we obtain another proof of normalization for the Barendregt–Coppo–Dezani intersection type assignment system.

Model-based deep embedding for constrained clustering analysis of single cell RNA-seq data

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.

Development of a Typology of Diabetes-Specific Family Functioning Among Adults With Type 2

Background Family members’ responses to adults’ diabetes and efforts to manage it vary widely. Multiple aspects of diabetes-specific family functioning have been identified as important for self-management and psychosocial well-being in theoretical (i.e., theories of social support and collaborative coping) and observational literature. Purpose Develop a typological framework of diabetes-specific family functioning and examine cross-sectional associations between type and diabetes outcomes. Methods We used electronic health record (EHR) data to identify a cohort of 5,545 adults receiving outpatient care for type 2 diabetes and invited them to complete a survey assessing 10 dimensions of diabetes-specific family functioning. We used k-means cluster analysis to identify types. After type assignment, we used EHR data for the full cohort to generate sampling weights to correct for imbalance between participants and non-participants. We used weighted data to examine unadjusted associations between participant characteristics and type, and in regression models to examine associations between type and diabetes outcomes. Regression models were adjusted for sociodemographics, diabetes duration, and insulin status. Results We identified and named four types: Collaborative and Helpful (33.8%), Satisfied with Low Involvement (22.2%), Want More Involvement (29.6%), and Critically Involved (14.5%; reflecting the highest levels of criticism and harmful involvement). Across these types, hemoglobin A1c, diabetes distress, depressive symptoms, diabetes medication adherence, and diabetes self-efficacy worsened. After covariate adjustment, type remained independently associated with each diabetes outcome (all p’s < .05). Conclusions The typology extends theories of family support in diabetes and applications of the typology may lead to breakthroughs in intervention design, tailoring, and evaluation.

Species Diversity, Mating Strategy and Pathogenicity of Calonectria Species from Diseased Leaves and Soils in the Eucalyptus Plantation in Southern China

Many Calonectria species are causal agents of diseases on several forestry, agricultural and horticultural crops. Calonectria leaf blight is one of the most important diseases associated with Eucalyptus plantations and nurseries in Asia and South America. Recently, symptoms of leaf rot and leaf blight caused by Calonectria species were observed in a one-year-old Eucalyptus experimental plantation in GuangXi Province, southern China. To better understand the species diversity, mating strategy and pathogenicity of Calonectria species isolated from diseased tissues and soils, diseased leaves and soils under the trees from ten Eucalyptus urophylla hybrid genotypes were collected. Three hundred and sixty-eight Calonectria isolates were obtained from diseased Eucalyptus leaves and soils under these trees, and 245 representative isolates were selected based on the sampling substrates and Eucalyptus genotypes and identified by DNA sequence analyses based on the translation elongation factor 1-alpha (tef1), β-tubulin (tub2), calmodulin (cmdA) and histone H3 (his3) gene regions, as well as a combination of morphological characteristics. These isolates were identified as Calonectria hongkongensis (50.2%), C. pseudoreteaudii (47.4%), C. aconidialis (1.6%), C. reteaudii (0.4%) and C. auriculiformis (0.4%). This is the first report of C. reteaudii and C. auriculiformis occurrence in China. Calonectria pseudoreteaudii was isolated from both Eucalyptus diseased leaves and soils; the other four species were only obtained from soils. MAT1-1-1 and MAT1-2-1 gene amplification and mating type assignment results showed that C. pseudoreteaudii is heterothallic and an asexual cycle represents the primary reproductive mode, C. reteaudii and C. auriculiformis are likely to be heterothallic and C. hongkongensis and C. aconidialis are homothallic. Based on the genetic diversity comparisons for C. pseudoreteaudii isolates from diseased leaves and soils, we hypothesize that C. pseudoreteaudii in soils was spread from diseased leaves. Both the mycelia plug and conidia suspension inoculations indicated that all five Calonectria species were pathogenic to the two Eucalyptus genotypes tested and the tolerance of the two genotypes differed. It is necessary to understand the ecological niche and epidemiological characteristics of these Calonectria species and to select disease resistant Eucalyptus genotypes in southern China in the future.

Kripke-style Semantics and Completeness for Full Simply Typed Lambda Calculus

Full simply typed lambda calculus is the simply typed lambda calculus extended with product types and sum types. We propose a Kripke-style semantics for full simply typed lambda calculus. We then prove soundness and completeness of type assignment in full simply typed lambda calculus with respect to the proposed semantics. The key point in the proof of completeness is the notion of a canonical model.

COMPLEXITY OF THE INFINITARY LAMBEK CALCULUS WITH KLEENE STAR

We consider the Lambek calculus, or noncommutative multiplicative intuitionistic linear logic, extended with iteration, or Kleene star, axiomatised by means of an $\omega $ -rule, and prove that the derivability problem in this calculus is $\Pi _1^0$ -hard. This solves a problem left open by Buszkowski (2007), who obtained the same complexity bound for infinitary action logic, which additionally includes additive conjunction and disjunction. As a by-product, we prove that any context-free language without the empty word can be generated by a Lambek grammar with unique type assignment, without Lambek’s nonemptiness restriction imposed (cf. Safiullin, 2007).