AI drug discovery in the peptide space has moved past in-silico screening into actual IND candidates. Coverage on this site spans Profluent's recombinase and AlphaGen work, MeddenoVo and Mexa-AI de novo platforms, Generate Biomedicines, and the HMD-AMP transformer-based AMP discovery platform from HLB Innovation (Nature Biomedical Engineering).
The most cited 2026 reads: AACR 2026 abstracts on AI-discovered cyclic peptides advancing toward IND, Profluent's gene-editing recombinase work that opens a new functional design axis, and the DDA-BERT transformer architecture for peptide identification. STAT News' BIO 2026 closing coverage (Alex Hogan, June 26) framed the year's organizing thesis: AI drug discovery moved 'from investment thesis to clinical fact,' anchored on Insilico Medicine's Rentosertib (TNIK inhibitor for IPF, Nature Medicine 2025) as the first peer-reviewed Phase IIa for a fully AI-discovered drug. The 2026 deal flow that priced that thesis: Chai-Lilly (Jan), Unnatural Products-Novartis $1.7B (Feb), Pinnacle Series B $89M (Mar), Parabilis-Regeneron $2.3B + $670M record IPO (May-June), Insilico-SK Biopharm $2.5B (June 22 BIO opening), Biogen-Dayra $50M (Nov 2025), and MindRank AI's $52M Series B on July 9, 2026 to push AI-designed oral GLP-1RA MDR-001 through the Phase 3 MOBILE trial (~750 participants) in China at cumulative R&D spend of $23M from concept to Phase 3.
Stories here cover platform launches, validation papers, and partnership deals. See #ai, #drug-discovery, and #peptide-discovery for adjacent threads.
Receptor.AI and Sethera Therapeutics announced Monday August 3, 2026 an integrated discovery alliance for AI-guided polymacrocyclic peptide medicines. Under the collaboration: Sethera generates and experimentally screens architecture-diverse polymacrocyclic peptide libraries; Receptor.AI applies physics-based modeling, artificial intelligence, and multiparameter optimization to interpret sequence, architecture, enrichment, and activity data and guide focused optimization cycles. The companies then design, synthesize, and experimentally test new candidates, feeding the resulting data into each subsequent cycle in a closed-loop discovery workflow. Sethera's platform installs 1-6 stable cross-links to generate polymacrocyclic, nested, in-line, and interpeptide structures across large encoded libraries, and (unlike constrained-peptide approaches centered on a predetermined structural motif) explores multiple experimentally accessible topologies. The collaboration targets historically hard-to-drug therapeutic targets that resist small-molecule and antibody approaches. The macrocyclic peptide modality has been commercially validated in 2026 through Merck's LIPFENDRA (enlicitide, FDA-approved July 16, 2026) as the first once-daily oral macrocyclic peptide PCSK9 inhibitor with $315/month launch pricing; Sethera and Receptor.AI aim to extend the modality to additional receptor targets.
MindRank AI, a Chinese clinical-stage biotech built around a proprietary Molecule Arts Platform (MAP) integrating biology, chemistry, computation, experimental evidence, and clinical learning, announced Thursday July 9, 2026 the completion of a $52 million Series B financing led by a group of institutional and healthcare funds. The company's lead program, MDR-001, is an AI-designed oral small-molecule GLP-1 receptor agonist that entered Phase 3 development in China in 2025 with the initiation of the MOBILE Phase 3 trial enrolling approximately 750 participants with overweight or obesity. The trial evaluates 52-week efficacy and safety. MindRank reports cumulative R&D investment from project initiation through the start of Phase 3 in China of approximately $23 million, with the program advancing from concept to Phase 3 in roughly 4.5 years. The financing extends the oral-GLP-1 competitive set beyond Eli Lilly's Foundayo (orforglipron), Novo Nordisk's Wegovy pill (oral semaglutide 25 mg), and Structure Therapeutics' aleniglipron. Anticipated commercial launch: within two to three years.
STAT News' closing coverage of BIO 2026 in San Diego across two pieces, Alex Hogan's June 26 STATus Report and Damian Garde's June 25 key-takeaways feature, identified two organizing themes that emerged from the four-day convention. First, AI drug discovery shifted from investment-thesis territory to clinical-fact territory: Insilico Medicine's rentosertib (a TNIK inhibitor for idiopathic pulmonary fibrosis where both the target and the compound were identified by generative AI, not a human chemist) became the first peer-reviewed Phase IIa result for a fully AI-discovered drug when results were published in Nature Medicine in 2025 (60 mg once daily produced +98.4 mL mean FVC change versus -20.3 mL placebo across 71 patients). Second, biotech executives at BIO 2026 expressed structural anxiety about Chinese drug development that the industry's primary legislative response, the BIOSECURE Act, does not address: the concern is about the speed and depth of Chinese scientific output rather than about narrow IP or supply-chain risks. The convention's mood reportedly shifted notably from the depressed atmosphere at industry events in 2024-2025 (when capital markets were closed and FDA reviewer turnover was concerning) toward a more constructive engagement with the new operational reality.
The BIO International Convention opens Monday June 22 in San Diego through Thursday June 25, drawing 20,000+ industry leaders across 130+ sessions in 18 focus areas — the year's largest peptide-and-biopharma deal-making and pipeline conference. Among the opening-day announcements: Insilico Medicine and SK Biopharmaceuticals will formally announce a research and development collaboration to discover AI-enabled innovative drug candidates in the neuroimmune CNS area, with Insilico eligible for up to $18M in upfront and near-term milestone payments and total potential deal value exceeding $2.5B including development, regulatory, and commercial milestones plus single-digit royalties. The pact sets a record for Insilico's APAC partnership totals. Industry watchers will track BIO 2026's deal-making cadence as the post-ADA / post-ENDO cycle concludes and the FDA PCAC peptide-compounding decision (July 23-24) becomes the next major regulatory catalyst.
Profluent, the Bezos-backed AI biotech, announced April 28 a multi-program research collaboration with Eli Lilly worth up to $2.25B in development and commercial milestones plus tiered royalties. Profluent will use its AI platform to design custom recombinases that integrate or modify large stretches of DNA at precise genomic loci, with the goal of generating site-specific gene-editing therapies for diseases of severe unmet need. The deal extends Lilly's diversification beyond peptide therapeutics, following a $1.12B Seamless Therapeutics recombinase deal in January 2026 and the closed Kelonia Therapeutics, Centessa, Orna, and Ajax acquisitions earlier in the quarter.
Daiichi Sankyo's collaboration with Meddenovo's AI-powered Mexa technology — for de novo design of cyclic peptides — is among the highlighted partnerships at the Boston peptide summit. The deal positions Daiichi to leverage Meddenovo's machine-learning platform for peptide candidates beyond the company's traditional small-molecule oncology focus. Cyclic peptide AI design is one of the fastest-moving subcategories in peptide drug discovery, with Circle Pharma, Bicycle Therapeutics, and Unnatural Products all pursuing distinct platform architectures.
Novo Nordisk and OpenAI announced a partnership to deploy AI across R&D, manufacturing, and corporate functions, with pilot programs launching immediately and full integration targeted by end of 2026. The deal positions Novo to analyze complex datasets, identify new drug candidates, and compress R&D timelines as it fights to claw back market share from Eli Lilly. OpenAI will also provide AI literacy training to Novo's global workforce.