Browse our latest Computational and Systems Biology articles

A broadly applicable deep-learning based method that recognizes rod-shaped bacteria of virtually any species in complex bacterial communities.

Laboratory experiments using a microbial community and mathematical modeling reveal how environmental disturbances can predictably alter the diversity and composition of an ecosystem.

Funding inequalities for NIH-supported research project grant investigators and organizations have increased over the past 25 years, but have modestly decreased over the past 3 years.

PARROT makes it easy for anyone to train and use system- or data-specific deep learning models that map between protein sequence and arbitrary sequence annotations.

Mobile phone data reveal aspects of human mobility patterns in Sub-Saharan Africa missed by standard spatial models however, model estimates can be improved by accounting for trip urbanicity and region.

Chromatin/transcriptome profiling in multiple mouse models with mutations in epigenetic machinery (EM) reveals shared abnormalities including many IgA-relevant genes, indicating that this kind of joint analysis may elucidate the multigenic nature of EM disorders.

A neural network model with asymmetric synaptic interactions can store and retrieve multiple continuous memory sequences with their temporal structure.

The person-specific adjusted HbA_1c addresses non-glycaemic variation in laboratory HbA_1c due to red blood cell lifespan differences, potentially providing a better marker to predict diabetes complications and guide glycaemic management.

Minimal modifications of common regulatory circuits can allow bacteria to learn from the past to better predict the future, on a physiological, rather than evolutionary, timescale.

The factors which drive and control disease progression can be inferred from mathematical models that integrate measures of immune responses, data from tissue sampling and markers of infection dynamics.