Prediction of organic compound aqueous solubility using machine learning: a comparison study of descriptor-based and fingerprints-based models.

A reliable and practical determination of a chemical species' solubility in water continues to be examined using empirical observations and exhaustive experimental studies alone. Predictions of chemical solubility in water using data-driven algorithms can allow us to create a rationally designed, efficient, and cost-effective tool for next-generation materials and chemical formulations. We present results from two machine learning (ML) modeling studies to adequately predict various species' solubility using data for over 8400 compounds. Molecular-descriptors, the most used method in previous studies, and Morgan fingerprint, a circular-based hash of the molecules' structures, were applied to produce water solubility estimates. We trained all models on 80% of the total datasets using the Random Forest (RFs) technique as the regressor and tested the prediction performance using the remaining 20%, resulting in coefficient of determination (R2) test values of 0.88 and 0.81 and root-mean-square deviation (RMSE) test values 0.64 and 0.80 for the descriptors and circular fingerprint methods, respectively. We interpreted the produced ML models and reported the most effective features for aqueous solubility measures using the Shapley Additive exPlanations (SHAP) and thermodynamic analysis. Low error, ability to investigate the molecular-level interactions, and compatibility with thermodynamic quantities made the fingerprint method a distinct model compared to other available computational tools. However, it is worth emphasizing that physicochemical descriptor model outperformed the fingerprint model in achieving better predictive accuracy for the given test set.

Examining the relations between spatial skills and mathematical performance: A meta-analysis.

Much recent research has focused on the relation between spatial skills and mathematical skills, which has resulted in widely reported links between these two skill sets. However, the magnitude of this relation is unclear. Furthermore, it is of interest whether this relation differs in size based on key demographic variables, such as gender and grade-level, and the extent to which this relation can be accounted for by shared domain-general reasoning skills across the two domains. Here we present the results of two meta-analytic studies synthesizing the findings from 45 articles to identify the magnitude of the relation, as well as potential moderators and mediators. The first meta-analysis employed correlated and hierarchical effects meta-regression models to examine the magnitude of the relation between spatial and mathematical skills, and to understand the effect of gender and grade-level on the association. The second meta-analysis employed meta-analytic structural equation modeling to determine how domain-general reasoning skills, specifically fluid reasoning and verbal skills, influence the relationship. Results revealed a positive moderate association between spatial and mathematical skills (r = .36, robust standard error = 0.035, τ2 = 0.039). However, no significant effect of gender or grade-level on the association was found. Additionally, we found that fluid reasoning and verbal skills mediated the relationship between spatial skills and mathematical skills, but a unique relation between the spatial and mathematical skills remained. Implications of these findings include advancing our understanding for how to leverage and bolster students' spatial skills as a mechanism for improving mathematical outcomes.

New geographic records for Echinococcus canadensis in coyotes and moose from Nova Scotia, Canada.

Echinococcus spp. tapeworms can cause serious diseases in mammals, including humans. Within the E. granulosus species complex, metacestodes produce unilocular cysts that are responsible for cystic echinococcosis in animal intermediate hosts. Canids are definitive hosts, harbouring adult cestodes in their intestines. Adult E. canadensis were recovered from the small intestine of 1 of 262 coyotes (Canis latrans) from Nova Scotia, Canada. Subsequently, we found unilocular cysts in lungs and livers of 4 of 8 sympatric moose (Alces alces) from Cape Breton Island. DNA was extracted from three cysts using the Qiagen DNeasy Blood and Tissue kit and assayed by polymerase chain reaction (PCR) with primers (cest4 and cest5) for a 117-bp region of the small subunit of ribosomal RNA of E. granulosus sensu lato, and further validated as E. canadensis G8 using primers targeting nicotinamide adenosine dinucleotide dehydrogenase subunit 1 (ND1) and cytochrome c oxidase subunit 1 (CO1) mitochondrial genes. These are the first records of E. canadensis in any of the three Maritime provinces, which include Nova Scotia, New Brunswick, and Prince Edward Island. The parasite was thought to be absent in this region due to extirpation of wolves (Canis spp.) in the 1800s. These findings suggest that further wildlife surveillance and risk assessment is warranted.