JEWJITSU PEPTIDES Research Roundup: GPCR peptide design, symbiont-regulated immunity, and PMOS metabolomics

Assessment of generative de novo peptide design methods for GPCRs reports a two‑part computational benchmark. The authors simulated validation of 91 GPCR–peptide and 22 GPCR–protein complexes using three prediction pipelines (AlphaFold2 Initial Guess, Boltz‑2, RosettaFold3) and then tested generative methods (BindCraft, BoltzGen, RFdiffusion3) for peptide sampling. The study highlights that current pipelines often overestimate confidence for peptides that are misplaced during validation, and it documents occurrences of memorization in both prediction and generation workflows. These observations address the gap between computational confidence metrics and experimental success.

The symbiont–host study examines Dendroctonus valens in association with its mutualistic fungus Leptographium procerum and the antagonist Ophiostoma minus. The authors report that L. procerum induces host HSP83 without triggering antimicrobial peptide (AMP) expression, whereas O. minus activates pattern recognition receptors including PGRP‑SA, PGRP‑SC2, βGRP3 and βGRP5. HSP83 is described as engaging a two‑tiered negative regulation: associating with several PRRs to attenuate detection of the antagonist and interacting with an NF‑κB‑like factor Dorsal to limit AMP production. The study frames this as a symbiont‑induced, mutation‑independent mechanism that may support selective defence during combined exposures.

The metabolomics analysis of polyendocrine metabolic ovarian syndrome (PMOS) uses a retrospective cross‑sectional cohort of 4,768 premenopausal patients from a biobank (samples 05/2010–08/2023). Within this set, PMOS cases numbered 403 and non‑PMOS 4,365. A panel of 71 metabolomic parameters spanning lipids, amino acids and glycolysis‑related metabolites was analysed with adjustments for age, BMI, exercise, metformin, OCPs and GLP‑1 agonists. Overall, PMOS status associated with 1/71 parameters in the full sample; in normal BMI there were no differences except lower creatinine, in overweight BMI 4/71 differences (mostly amino acids) were reported, and in obesity there were no associations. The study notes that metabolomic signatures vary by BMI phenotype.