What successful incretin drugs, BPC-157, and Epitalon teach us about peptide engineering, translational evidence, and the limits of molecular plausibility.
A concise overview of how molecules and chemical knowledge are represented for machines and humans alike, from SMILES strings to LLM-based agents.
What agentic coding changes for computational drug discovery: reliability, the shift of the bottleneck onto validation, and where domain judgment still decides the science.
Computational reproduction of the C-Se reductive elimination step in gold(III)-catalyzed selenocysteine arylation using ORCA, GFN2-xTB, and OpenBabel.
The shift from prompt-centric AI usage to full-stack systems engineering, covering context engineering, agentic loops, retrieval pipelines, evaluation frameworks, and cost-aware deployment.
What CoMFA and CoMSIA actually compute, how their fields differ, and what that means for alignment, PLS models and contour maps.
How to move from textbook bioisosteres to substituent property spaces, matched molecular pairs and SwissBioisostere, and why every suggestion is still a hypothesis.
PCA, t-SNE, and UMAP can produce very different views of the same molecular dataset. The reason is simple: chemical space depends on how molecules are represented, how similarity is defined, and how that high-dimensional information is projected.
How metastable states turn fragmented molecular dynamics trajectories into kinetic models—and how to distinguish meaningful molecular states from attractive but misleading clusters.
A practical map of how enhanced-sampling and rare-event methods reveal molecular pathways, free-energy landscapes, and kinetics—and what each method can legitimately tell us.