Domain-informed graph neural networks: A quantum chemistry case study.

We explore different strategies to integrate prior domain knowledge into the design of graph neural networks (GNN). Our study is supported by a use-case of estimating the potential energy of chemical systems (molecules and crystals) represented as graphs. We integrate two elements of domain knowledge into the design of the GNN to constrain and regularise its learning, towards higher accuracy and generalisation. First, knowledge on the existence of different types of relations/graph edges (e.g. chemical bonds in our case study) between nodes of the graph is used to modulate their interactions. We formulate and compare two strategies, namely specialised message production and specialised update of internal states. Second, knowledge of the relevance of some physical quantities is used to constrain the learnt features towards a higher physical relevance using a simple multi-task learning (MTL) paradigm. We explore the potential of MTL to better capture the underlying mechanisms behind the studied phenomenon. We demonstrate the general applicability of our two knowledge integrations by applying them to three architectures that rely on different mechanisms to propagate information between nodes and to update node states. Our implementations are made publicly available. To support these experiments, we release three new datasets of out-of-equilibrium molecules and crystals of various complexities.

Availability of Naloxone in 2 Underserved Urban Communities in Georgia.

In 2016, unintentional injuries became the third leading cause of death in the United States. In 2018, 54% of 103 672 unintentional injury deaths were due to drug overdoses among adults 19 to 64 years of age. In Georgia, opioid overdose deaths continued to increase, despite a 2014 state law for naloxone use to prevent deaths, and a 2017 amendment for more widespread community use without a prescription. Given these policies, naloxone availability in pharmacies in underserved communities remains unclear. Our objective is to explore naloxone availability in such communities. Three Public Health and Preventive Medicine residents during a social-cultural-behavioral longitudinal rotation conducted interviews of 9 community pharmacists. Several themes emerged: more education was needed, and naloxone was available only by prescription in certain pharmacies or in limited amounts. Additional assessments among community members and sectors can examine the extent to which policies to expand naloxone availability and accessibility are implemented, including reduced naloxone costs.

Dopamine-2 receptor activation suppresses PACAP expression in gonadotrophs.

Pituitary adenylate cyclase-activating polypeptide (PACAP) is expressed at a high level in the fetal pituitary and decreases profoundly between embryonic day 19 and postnatal day 1 (PN1), with a further decrease from PN1 to PN4. In this series of experiments, we investigated the hypothesis that dopamine 2 receptor (Drd2) activation interrupts a cAMP-dependent feed-forward loop that maintains PACAP expression at a high level in the fetal pituitary. Using single-cell RT-PCR of pituitary cell cultures from newborn rats, Drd2 mRNA was identified in gonadotrophs that were also positive for PACAP mRNA. PACAP expression in pituitary cultures from embryonic day 19 rats was suppressed by the PACAP6-38 antagonist and by the Drd2 agonist bromocriptine. Increasing concentrations of bromocriptine inhibited cAMP production as well as cAMP signaling based on cAMP response element-luciferase activity, decreased PACAP promoter activity, and decreased PACAP mRNA levels in αT3-1 gonadotroph cells. Furthermore, blockade of dopamine receptors by injecting haloperidol into newborn rat pups partially reversed the developmental decline in pituitary PACAP mRNA that occurs between PN1 and PN4. These results provide evidence that dopamine receptor signaling regulates PACAP expression under physiological conditions and lend support to the hypothesis that a rise in hypothalamic dopamine at birth abrogates cAMP signaling in fetal gonadotrophs to interrupt a feed-forward mechanism that maintains PACAP expression at a high level in the fetal pituitary. We propose that this perinatal decline in pituitary PACAP reduces pituitary follistatin which permits GnRH receptors and FSH-ß to increase to facilitate activation of the neonatal gonad.