Drug discovery in peptides is no longer just about phage display. Recent trends covered on this site: AI-designed peptides moving toward IND, biocatalysis using peptide asparaginyl ligase (PAL) for cyclization, transformer-based methods for peptide identification (DDA-BERT) and immunopeptidomics, and a growing number of de novo platforms (Profluent's AlphaGen, Generate Biomedicines, MeddenoVo, Mexa-AI, Cradle).
The macrocyclic and bicyclic peptide chemistry powering Bicycle Therapeutics, Circle Pharma, and Sapience Therapeutics sits here too. The undruggable targets — KRAS, beta-catenin, intracellular protein–protein interactions — are the discovery frontier that small-molecule chemistry has struggled to address.
Stories below cover platform launches, partnership deals, and validation papers. See #peptide-discovery and #ai-drug-discovery for narrower threads.
A Frontiers in Drug Discovery review (April 10, 2026) catalogs underexplored peptide-receptor systems that the authors argue failed not for biological reasons but because of technical and conceptual barriers solvable with modern peptide engineering. Coverage spans metabolic and energy-balance peptides (apelin, spexin), appetite-regulating systems (peptide YY, oxyntomodulin), bone-muscle-fat crosstalk mediators (osteocalcin, irisin), and neuroendocrine-immune-metabolic peptides (phoenixin, relaxin-3). The argument: lessons from GLP-1 — stabilization, conjugation, and dosing innovation — now make these orphan receptor systems tractable. Companion essay from Bloomgarden in the Journal of Diabetes (2026, vol 18 e70204) frames the same opportunity from the clinical side, citing GLP-1 tolerability ceilings and the 10-20% non-responder problem that creates room for the next wave.
A May 2026 Drug Discovery World feature consolidates the case for cyclic peptides as a distinct therapeutic modality: larger and more selective than small molecules, more permeable and cheaper to manufacture than antibodies, and uniquely suited for protein-protein-interaction targets that have resisted traditional drug discovery. The piece traces the recent algorithmic stack — RFDiffusion adaptations for cyclic backbones, AfCycDesign, ProteinMPNN-derived sequence design — alongside the synthesis chemistry advances (one-pot ligations, photo-redox macrocyclizations) that turn computational hits into tractable scaffolds. Over 40 cyclic peptide drugs are now FDA-approved across endocrine, oncology, and antimicrobial uses, and 6+ peptide-drug conjugates sit in Phase 3, per the late-April PDC market analysis.
An ACS Biochemistry paper published April 10 examined the maturation of AI-designed peptides as tools for biochemistry research and therapeutic development. The work covers computational design of cyclic peptides, antimicrobial peptides, and peptide ligands with novel binding specificities — capturing the moment when machine-learning-driven peptide design has begun delivering candidates competitive with traditional medicinal-chemistry approaches across multiple modality categories.
A Nature Communications paper introduced CycloSEL (Cyclic Self-Encoded Libraries), an end-to-end workflow that screens synthetic macrocycle libraries enriched in drug-like 'beyond rule of five' features using affinity selections and tandem mass spectrometry — eliminating the genetic-barcode requirement of traditional macrocyclic peptide discovery. The team validated the approach against the oncology target carbonic anhydrase IX with a 16-million-member library, achieving robust enrichment and accurate identification of true binders. The platform shifts peptide drug discovery toward small molecule-like drug-likeness optimization from day one.
University of Nebraska Medical Center's Guangshun Wang lab released APD6, the expanded antimicrobial peptide database, containing 6,309 peptides (3,379 natural AMPs, 2,290 synthetic, 373 AI-predicted) as of January 2026. New features include the Antimicrobial Peptide Information Pipeline (AMPIP) and expanded functional wheel covering anticancer and antidiabetic activity — positioning APD6 as the most comprehensive reference for AMP drug discovery as AI-assisted antibiotic design accelerates.
A Frontiers in Pharmacology review cataloged 176 neuropeptides across 16 families found in the venoms of 107 scorpion species, highlighting Buthus martensii Karsch (BmK) peptides as selective ion channel modulators. Examples include MarTX (selective for BK(α+β4) channels) and BmKTX (Kv1.3 blocker) — positioning scorpion venom as an underexplored source of next-generation neurotherapeutic leads for pain, epilepsy, and autoimmune conditions.
A comprehensive review in Discover Oncology highlights antimicrobial peptides' emerging dual role as anticancer and antiviral therapeutics. AMPs selectively target cancer cell membranes through electrostatic interactions while also demonstrating antiviral activity, with their immunomodulatory properties and reduced resistance development offering advantages over conventional chemotherapy.
A study in Nature Microbiology used a generative protein language model (ProteoGPT) to discover novel antimicrobial peptides effective against multidrug-resistant bacteria. The AI-designed peptides showed comparable or superior efficacy to clinical antibiotics in mouse infection models, with reduced resistance development and no organ damage.
A new phage display platform using ~482 venom-derived peptide scaffolds and machine learning-guided mutation design achieved a 100% success rate identifying strong binders across four diverse therapeutic targets (CD47, DLL3, IL33, P2X7R). The approach could accelerate venom peptide drug discovery for cancer, inflammation, and pain.
A study published in Nature Microbiology used a generative AI approach to discover antimicrobial peptides effective against multidrug-resistant bacteria. The deep learning pipeline screened millions of candidate sequences, achieving a 94.4% success rate in lab validation, with two candidates showing exceptional efficacy and low resistance potential.