Highlight of PDGrapher in @thecrimson nature.com/articles/s41551-0… Great work by @justguadaa Xiang Lin and collaboration with K. Veselkov and @mmbronstein

Ever wish you could hit "undo" on disease? 🩺🔄 nature.com/articles/s41551-0… Most drug discovery asks: what does this perturbation do to cells? But we can also ask the reverse: which perturbations undo a disease signature and move cells back toward health? That's the idea behind PDGrapher, just published in Nature Biomedical Engineering @natBME @justguadaa @mmbronstein ⚠️ Challenge Phenotype-based discovery screens compounds forward, measuring whether they improve disease phenotypes But it does not address which perturbations will reverse disease signatures or capture combinatorial targets that may be essential for restoring health 💡 Approach PDGrapher tackles perturbagen discovery: finding the genes, drugs, or combinations that can shift a diseased cell state toward health It represents disease cell states on molecular networks and asks: which perturbations will move these states toward health? It uses a causally inspired model that overlays gene-expression signatures onto proxy causal graphs. Perturbagens are encoded as edge modifications, allowing the model to reason about how interventions reshape cellular processes It is trained with two objectives: 💡 Cycle-consistency loss: ensures predicted perturbations move a diseased state toward its observed treated state 💡 Supervision loss: directly aligns predictions with experimental treatment outcomes

Congrats @LogmlSchool for running an in-person grad ML summer school at @imperialcollege and opening up research opportunities for students worldwide! 👏 logml.ai Thanks to organizers, @valegiunca, and mentors @justguadaa @_cagarwal @ruthie_johnson @YEktefaie @oq_35

On knowing a gene: Distributional hypothesis of gene function Just as words derive meanings from context, genes can switch their roles with biological surroundings. Traditional gene annotations miss this Advances in transformers suggest a new perspective: gene functions as distributions over cellular contexts, similar to how words have context-dependent meanings cell.com/cell-systems/fullte… @CellSystemsCP Genes can exhibit pleiotropy, meaning they have multiple functions based on their biological context. Current ontologies often fail to capture this complexity Recent NLP breakthroughs, where word semantics are automatically learned from diverse contexts, offer a valuable analogy for understanding gene functions Modern distributional semantics uses vectors in a learned semantic space, driving innovations in LLMs and GPT models We outline a similar paradigm shift for gene functions. Viewing them as distributions over cellular contexts could empower models trained from extensive biological datasets @Harvard @HarvardDBMI @KempnerInst @harvard_data @DanaFarber @broadinstitute @GoogleDeepMind @joshbiology @JKwonBio @justguadaa W. Hahn