Our Take
This is a competent partnership between a carrier (LG's AI models) and a domain specialist (D&D's peptide chemistry), not a capability breakthrough—the real test is whether the feedback loop between AI-generated candidates and wet-lab validation produces candidates faster or cheaper than existing methods.
Why it matters
Oral peptide delivery is a known hard problem in pharma; peptides degrade in the digestive system, forcing injection-only routes. If the partnership reduces the time or cost to identify candidates that survive oral administration, it narrows the gap between computational discovery and clinical viability.
Do this week
Biotech AI leads: map your current peptide candidate discovery cycle (time, cost, attrition rate) against LG and D&D's published results once preclinical data emerge, so you can decide whether this model applies to your pipeline.
LG AI and D&D Pharmatech formalize peptide drug partnership
LG AI Research and D&D Pharmatech signed a collaboration agreement to jointly discover and develop oral peptide therapeutics, announced at a ceremony at LG Twin Towers in Seoul. LG AI Research head Woohyung Lim and D&D Pharmatech CEO Seulki Lee attended the signing.
The project divides labor: LG AI Research will develop computational models to design and discover new peptide candidates by analyzing disease-causing substances and identifying optimal peptide sequences. D&D Pharmatech will handle structural design, synthesis, evaluation, formulation for oral delivery, and preclinical and clinical development.
The partnership operates on a feedback loop. LG AI provides AI-generated candidates; D&D validates them experimentally; results inform the next round of model improvements. The goal is to create tablet-form treatments that improve safety and absorption rates for peptides, which are normally administered by injection because they break down in the digestive system.
Oral peptide delivery is a known bottleneck in drug development
Peptides perform critical physiological functions but face a structural barrier: the digestive system degrades them before absorption. This forces them into injection-only routes, limiting patient compliance and market reach. Conventional structure-based design has not solved this problem at scale.
The partnership targets this friction point by pairing machine-learning-assisted design (rapid candidate generation and ranking) with wet-lab validation (proof that candidates actually work). If the feedback loop accelerates the discovery of orally bioavailable peptide sequences, it could compress the time from target identification to IND-enabling data.
The deal is notably narrow in scope: it does not claim to have solved oral peptide delivery, only to be working on it with AI assistance. LG AI Research is also developing other biotech AI platforms, including oncology analysis tools and the EXAONE Discovery platform for drug and material research, signaling broader intent in drug discovery acceleration.
Evaluate this model against your candidate pipeline
For biotech and pharma teams running peptide programs: the real measure is throughput and cost. Ask D&D and LG for preclinical data once available: How many AI-suggested candidates entered synthesis? What fraction showed oral activity? How did time-to-candidate and cost-per-candidate compare to your baseline (literature, prior internal efforts, or vendor alternatives)?
This partnership is not a product launch or a published benchmark. It is a two-company collaboration with no published results yet. Wait for preclinical validation before integrating into your own discovery roadmap.