Analysis of Quantitative Evaluation Methods for Synergistic Effects of Antengene's AI+TCE Platform

Based on the searched information, this article systematically analyzes the quantitative evaluation methods for the synergistic effects of Antengene’s AI R&D department on the layout of its TCE platform.

Antengene announced the establishment of an AI R&D department in early 2025, deploying DeepSeek technology to accelerate the layout of its second-generation T-cell engager (TCE) platform [1]. Its self-developed AnTenGager™ platform has a steric hindrance shielding effect, which can induce disease-associated antigen (DAA)-dependent T-cell binding and activation, achieving strong activity while reducing the risk of cytokine release syndrome (CRS) [2].

According to industry benchmark data, the use of AI technology can shorten the time required for early drug discovery to 1/3 of traditional methods [3]. For complex antibody drugs like TCE, AI can significantly shorten the cycle in the following links:

• Target discovery and validation
  : From months to weeks
• Molecular design and optimization
  : AI-assisted virtual screening can reduce experimental times by 60-70%
• Lead compound optimization
  : From dozens of iterations to 1/3 of the cycle with AI

Cost Item	Traditional Model	AI-Enabled Model	Savings Rate
Early Discovery Cost	Baseline 100%	~50%	~50%[3]
Preclinical Research	Baseline 100%	~60-70%	30-40%
Total R&D Cost	Baseline 100%	~30-50%	50-70%

Based on the key financial data disclosed in the company’s 2024 annual report, the initial results of the AI strategy layout can be observed [4][5]:

:

• R&D costs in 2024 decreased from RMB 406 million in 2023 to RMB 259 million, a year-on-year decrease of 36%
• Among them, “drug development expenses and R&D personnel costs decreased by RMB 97.3 million”
• The company clearly stated that this decrease “was mainly due to the improvement of R&D efficiency”

:
Antengene’s TCE pipeline has made steady progress, including:

• ATG-201 (CD19 x CD3 TCE)
• ATG-102 (LILRB4 x CD3 TCE)
• ATG-106 (CDH6 x CD3 TCE)
• ATG-107 (FLT3 x CD3 TCE)
• ATG-110 (LY6G6D x CD3 TCE)

• Pipeline advancement speed
  : Time compression ratio from PCC to IND
• Molecular optimization cycle
  : Number of iterations from first-in-class candidate to preclinical candidate
• Resource utilization rate
  : Number of pipelines output per R&D personnel

• Preclinical success rate
  : Proportion of AI-assisted designed molecules entering clinical trials
• Off-target risk reduction rate
  : Toxic events reduced through AI prediction
• Effectiveness improvement
  : Comparison of preclinical efficacy data with traditional methods

• R&D ROI
  : Pipeline value generated per unit of R&D investment
• Milestone achievement rate
  : Probability improvement of key nodes completed on time
• External cooperation value
  : Total cooperation orders with AI enterprises such as MindRank AI exceed 100 million US dollars [6]

The combination of the AnTenGager™ platform and AI technology is mainly reflected in the following synergistic mechanisms:

• AI-assisted molecular dynamics simulation to optimize steric hindrance shielding effect
• Deep learning to predict the optimal CD3 affinity balance point
• Computational simulation to reduce CRS risk

• Accurate prediction of disease-associated antigen-dependent activation characteristics
• AI optimization of correct folding and expression of bispecific antibodies
• Machine learning to accelerate optimal hinge region design

• AI prediction of immunogenicity risk
• Deep learning to evaluate cytokine release profile
• Intelligent screening to reduce clinical failure risk

:

• The synergistic effect of AI has a lag, which takes 6-12 months to appear in the pipeline progress
• Difficulties in quantitative attribution: R&D efficiency improvement is affected by multiple factors
• The industry lacks unified evaluation standards for AI drug R&D efficiency

:

1. Establish baseline comparison
   : Use R&D efficiency before the establishment of the AI department as the benchmark
2. Set milestone nodes
   : Track time changes of key nodes such as PCC achievement and IND application
3. Cost-benefit analysis
   : Compare R&D projects that did not use AI assistance in the same period
4. Success rate tracking
   : Statistically analyze the proportion change of AI-assisted designed molecules entering clinical trials

By establishing an AI R&D department and deploying DeepSeek technology, Antengene has injected significant synergistic potential into the AnTenGager™ TCE platform. Combining industry data and the company’s financial performance, its synergistic effects can be initially quantitatively evaluated:

• R&D cost reduction
  : Expected to achieve 30-50% cost savings
• Time efficiency improvement
  : R&D cycle can be compressed to 1/3 to 1/2 of traditional methods
• Success rate optimization
  : Improve clinical conversion rate through AI-assisted risk control

With the in-depth application of AI technology in TCE molecular design, Antengene is expected to establish a differentiated advantage in the highly competitive TCE track, and its technical characteristics of “synergy and toxicity reduction” will also be more fully verified.

---

:
[1] 2025中国创新药十大牛股
[2] 德琪医药宣布计划接入DeepSeek
[3] 潮声丨新老药企争相入局，AI做的药你敢吃吗？
[4] TCE 2.0蓄势待发德琪医药深度报告
[5] 德琪醫藥有限公司2024年度報告
[6] 北京医药行业协会信息周报