Developmental gene regulatory network atlas @GenomeBiology over 24 neurogenic lineages, a hierarchical roadmap revealing spatiotemporal dynamics across brain regions & species, and tracing evolution & disorder risk back to early mammalian neurogenesis, doi.org/10.1186/s13059…
Daifeng Wang
@daifengwang
Associate Professor @uwmadison @UWWaismanCenter. AI/ML to bridge computation and biology for mechanistic & translational insights into brains & diseases.
455 Following 385 Followers 408 Posts
Latest posts 5
Our knowledge-guided graph learning iBrainmap @panos_roussos @NatureComms infers personalized roadmap of each donor’s cell interactions & gene networks, revealing functional heterogeneity, disease subpopulations & trajectories across >1900 brains #PsychAD nature.com/articles/s4146…
Our AI framework PASCode w/@panos_roussos @NatureMedicine uses graph neural networks to score phenotypic association of >6M cells #PsychAD and discover unknown cell subtypes & genes in #Alzheimers for pathology, cognitive resilience & psychiatric symptoms nature.com/articles/s4159…
Our SegJointGene jointly segments cells and prioritizes spatial genes for cell-type-specific segmentation from spatial transcriptomics and proteomics data, using information-entropy-guided convolutional neural networks. academic.oup.com/bioinformatics…
academic.oup.com SegJointGene: joint cell segmentation and spatial gene prioritization by information entropy guided... AbstractMotivation. Spatial sequencing technologies enable the single-cell-level study of molecular organization in tissues. Revealing such spatial pattern CellTRIP infers virtual cell environment to simulate cell dynamics & interactions (beyond #virtualcell) by multi-agent reinforcement learning. It enables in-silico perturbation of any cells or genes, prediction of spatial/temporal changes, and gene prioritization for single-cells
Inferring virtual cell environments using multi-agent reinforcement learning biorxiv.org/content/10.110… #biorxiv_bioinfoView quoted post on X