Measuring HIV-1 DNA levels at the time of stopping antiretroviral therapy (when initiated during primary infection) predicts clinical progression and the time taken for plasma viraemia to become detectable.

Risk factors for SARS-CoV-2 infection in healthcare workers included caring for Covid-19 patients, Black or Asian ethnicity, and Covid-19-positive household contacts, whereas a bundle of PPE-related measures protected ICU staff.

Numerous related sample characteristics affect polygenic score performance, and incorporation of interaction effects increases both model and polygenic score performance.

The two most commonly used chemogenetic actuators, clozapine-N-oxide and compound 21, similarly modulate sleep in wild-type laboratory mice indicating that even designer drugs which do not convert to clozapine can elicit behavioural effects and require adequate controls.

An inference paradigm for extracting neuronal correlations from two-photon imaging data, without requiring intermediate spike deconvolution, provides significant performance gains over existing methods as demonstrated by theoretical analysis, simulation studies, and real-data applications.

Computational analysis reveals that additive summary statistics from genome-wide association studies can provide evidence of non-additive heritability in complex traits.

Combining computational models with FMRI in humans during pain reversal learning reveals parallel pain modulatory brain circuits that differ in their flexibility and sensitivity to instructions.

The lysine acetyltransferases KAT2A and KAT2B regulate the dynamic chromatin association of the tumour suppressor protein PALB2, protecting active genes and promoting DNA damage repair in a context-dependent manner.