Massimiliano Versace·1h agoBrains are “expectation machines”.What the cortex teaches us about machines that never stop learning, and why Physical AI needs it now
Devanik Debnath·8h agoThe Non-Abelian Geometry of Morphogenesis: When Order Becomes CurvatureA mathematical account of intervention order as holonomy, with an honest accounting of what the mathematics does and does not show
InLevel Up CodingbyNagesh Singh Chauhan·19h agoDemystify DNA Sequencing with AIHow pretrained genomic models learn biological patterns from massive DNA datasets and adapt their knowledge to specialized tasks such as…
Devanik Debnath·Nov 3How Many Antibodies Are Enough? The Mathematics of Coverage in Antigenic SpaceAn epsilon-net perspective on broad-spectrum antibody selection, VC dimension, and the mathematical cost of escape resistance.
InData Science CollectivebyMarko Mihajlovic·Oct 28Simulating DNA Replication in Python: Where Computer Science Meets Molecular BiologyFrom parallel replication forks to error correction and stochastic modeling, exploring what it takes to build a biologically meaningful…
Devanik Debnath·Oct 24Aging May Depend on When We Repair Damage, Not Just How MuchA computational thought experiment explores how repair order, interacting damage, and limited maintenance capacity can shape a theoretical…
Scufieldidiakhoa·Oct 22My AI Tried to Predict Cancer Gene Dependencies. A Simple Baseline Beat It.I spent more than a year building an AI system to investigate biological evidence. Then I tested one of its predictions against future…A response icon4A response icon4
Sivarat Malapet·Oct 21A fly-connectome simulation integrated with a live Doom game-engine arena, while also performing…My latest personal research project explores how a biological connectome can be transformed into a trainable artificial neural network.
Jaivardhan Samadhiya·Oct 20Decoding the Bacterial Armor: How Machine Learning Unlocks Phage Tailspike Proteins and…When we look at the cutting edge of computational phage therapy, a lot of attention goes to high-level pipelines and CRISPR editing. But…
