National Institutes of Health (NIH) grants have evolved over time, with a greater proportion of funds going to solicited projects, but inflation-adjusted costs have remained relatively stable following a brief increase during the NIH-doubling (1998-2003).

A longitudinal study of open access levels at 1,207 research institutions worldwide reveals high performing institutions in Latin America, Africa and Asia.

An ellipse of insignificant analysis is a robust method for ascertaining the strength of even large dichotomous outcome trials in biomedical science, and is a novel means to detect potentially dubious results and research fraud.

The Reproducibility Project: Cancer Biology will generate a high-quality dataset to explore questions about the reproducibility of research, and will make all data, analysis and other research materials openly available to the research community.

Using multiple analysts to independently analyze the same dataset has the potential to explore and strengthen the robustness of results and conclusions in basic and applied research.

The best papers provide evidence that can be used to make changes that improve the health and lives of people around the world.

Compiling public datasets into a single, centralised repository and linking directly to analytical software completely transforms the scale and scope of causal inference across the phenome.

White matter microstructure alterations in key bundles affected in Alzheimer's disease are related to amyloid and tau pathology in the preclinical phase of sporadic and autosomal-dominant Alzheimer's disease.

Epidemiological and viral sequence data can be combined to provide standardised feedback on cases of hospital onset COVID-19 infection that would be useful for infection prevention and control.

Open research practices bring significant benefits to researchers.