PhenoID, a language model normalizer of physical examinations from genetics clinical notes.

Background: Phenotypes identified during dysmorphology physical examinations are critical to genetic diagnosis and nearly universally documented as free-text in the electronic health record (EHR). Variation in how phenotypes are recorded in free-text makes large-scale computational analysis extremely challenging. Existing natural language processing (NLP) approaches to address phenotype extraction are trained largely on the biomedical literature or on case vignettes rather than actual EHR data. Methods: We implemented a tailored system at the Children's Hospital of Philadelpia that allows clinicians to document dysmorphology physical exam findings. From the underlying data, we manually annotated a corpus of 3136 organ system observations using the Human Phenotype Ontology (HPO). We provide this corpus publicly. We trained a transformer based NLP system to identify HPO terms from exam observations. The pipeline includes an extractor, which identifies tokens in the sentence expected to contain an HPO term, and a normalizer, which uses those tokens together with the original observation to determine the specific term mentioned. Findings: We find that our labeler and normalizer NLP pipeline, which we call PhenoID, achieves state-of-the-art performance for the dysmorphology physical exam phenotype extraction task. PhenoID's performance on the test set was 0.717, compared to the nearest baseline system (Pheno-Tagger) performance of 0.633. An analysis of our system's normalization errors shows possible imperfections in the HPO terminology itself but also reveals a lack of semantic understanding by our transformer models. Interpretation: Transformers-based NLP models are a promising approach to genetic phenotype extraction and, with recent development of larger pre-trained causal language models, may improve semantic understanding in the future. We believe our results also have direct applicability to more general extraction of medical signs and symptoms. Funding: US National Institutes of Health.

Bidirectional electromagnetic control of the hypothalamus regulates feeding and metabolism.

Targeted, temporally regulated neural modulation is invaluable in determining the physiological roles of specific neural populations or circuits. Here we describe a system for non-invasive, temporal activation or inhibition of neuronal activity in vivo and its use to study central nervous system control of glucose homeostasis and feeding in mice. We are able to induce neuronal activation remotely using radio waves or magnetic fields via Cre-dependent expression of a GFP-tagged ferritin fusion protein tethered to the cation-conducting transient receptor potential vanilloid 1 (TRPV1) by a camelid anti-GFP antibody (anti-GFP-TRPV1). Neuronal inhibition via the same stimuli is achieved by mutating the TRPV1 pore, rendering the channel chloride-permeable. These constructs were targeted to glucose-sensing neurons in the ventromedial hypothalamus in glucokinase-Cre mice, which express Cre in glucose-sensing neurons. Acute activation of glucose-sensing neurons in this region increases plasma glucose and glucagon, lowers insulin levels and stimulates feeding, while inhibition reduces blood glucose, raises insulin levels and suppresses feeding. These results suggest that pancreatic hormones function as an effector mechanism of central nervous system circuits controlling blood glucose and behaviour. The method we employ obviates the need for permanent implants and could potentially be applied to study other neural processes or used to regulate other, even dispersed, cell types.

A critical role for mTORC1 in erythropoiesis and anemia.

Red blood cells (RBC) must coordinate their rate of growth and proliferation with the availability of nutrients, such as iron, but the signaling mechanisms that link the nutritional state to RBC growth are incompletely understood. We performed a screen for cell types that have high levels of signaling through mTORC1, a protein kinase that couples nutrient availability to cell growth. This screen revealed that reticulocytes show high levels of phosphorylated ribosomal protein S6, a downstream target of mTORC1. We found that mTORC1 activity in RBCs is regulated by dietary iron and that genetic activation or inhibition of mTORC1 results in macrocytic or microcytic anemia, respectively. Finally, ATP competitive mTOR inhibitors reduced RBC proliferation and were lethal after treatment with phenylhydrazine, an inducer of hemolysis. These results identify the mTORC1 pathway as a critical regulator of RBC growth and proliferation and establish that perturbations in this pathway result in anemia.

Developmental role for endocannabinoid signaling in regulating glucose metabolism and growth.

Treatment of ob/ob (obese) mice with a cannabinoid receptor 1 (Cnr1) antagonist reduces food intake, suggesting a role for endocannabinoid signaling in leptin action. We further evaluated the role of endocannabinoid signaling by analyzing the phenotype of Cnr1 knockout ob/ob mice. Double mutant animals show a more severe growth retardation than ob/ob mice with similar levels of adiposity and reduced IGF-I levels without alterations of growth hormone (GH) levels. The double mutant mice are also significantly more glucose intolerant than ob/ob mice. This is in contrast to treatment of ob/ob mice with a Cnr1 antagonist that had no effect on glucose metabolism, suggesting a possible requirement for endocannabinoid signaling during development for normal glucose homeostasis. Double mutant animals also showed similar leptin sensitivity as ob/ob mice, suggesting that there are developmental changes that compensate for the loss of Cnr1 signaling. These data establish a role for Cnr1 during development and suggest that compensatory changes during development may mitigate the requirement for Cnr1 in mediating the effects of leptin. The data also suggest a developmental role for Cnr1 to promote growth, regulate the GH/IGF-I axis, and improve ß-cell function and glucose homeostasis in the setting of leptin deficiency.