NNewsGPT ← Home
CN

Huayuan Zhiyin Secures Millions in Seed Funding for AI Drug Efficacy Prediction

CN2 hr ago

Huayuan Zhiyin, a company specializing in AI virtual cells (AIVC), has successfully completed a seed funding round, raising tens of millions of RMB. The investment, led by Shuming Ventures, will fuel advancements in multimodal sequencing technology, expand collaborations with top-tier hospitals, and support team growth. The company is already planning its next funding round.

The founding team comprises experienced pharmaceutical industry professionals and computational biology researchers, augmented by a scientific advisory board including experts from the Shenzhen National Gene Bank. CEO Du Runshi is a serial AI entrepreneur, and CTO Wang Yixuan, a PhD candidate at the Chinese University of Hong Kong, previously developed a gene perturbation prediction model. Li Yu, an assistant professor at CUHK, serves as Chief AI Scientist.

Huayuan Zhiyin addresses the high failure rate in new drug development, particularly the gap between animal testing and human trials. Their core innovation, the Wise-Perturb application model, focuses on predicting drug efficacy in humans. Unlike traditional AI drug discovery that optimizes early-stage target identification, Wise-Perturb aims to determine a drug's true potential for human trials by simulating its effects at the cellular level. This involves integrating DNA, RNA, and protein data from single cells to create a comprehensive understanding of cellular responses to drugs.

The company's approach overcomes limitations of traditional models that often rely on simplified cell lines and lack patient-specific genetic information. Huayuan Zhiyin builds a three-tiered data foundation, incorporating general single-cell sequencing, extensive in-vitro cell line perturbation data, and scarce clinical trial data. They are establishing joint laboratories with leading hospitals to collect specific patient data for training specialized disease models, aiming for deep collaboration with at least 30 top hospitals within three years.

Wise-Perturb's cell-specific design allows for accurate predictions across different cell types and cancers without extensive retraining for each new scenario. Validation cases include predicting the efficacy of the ADC drug DS-8201 across different cancer types and stratifying patient response to Osimertinib for non-small cell lung cancer, with results aligning with clinical observations. This validated predictive capability enables Huayuan Zhiyin to engage in paid collaborations with medical institutions and pharmaceutical companies from an early stage.

Commercial services include preclinical pipeline evaluation and patient stratification for clinical trials for pharmaceutical companies, as well as joint research projects with hospitals on drug repurposing and target discovery. Future plans include co-development models with shared risk and revenue. Huayuan Zhiyin prioritizes small-scale validation projects to build trust through reproducible data, capitalizing on the industry's demand for accurate human efficacy prediction and the scarcity of mature alternative solutions.

AI Analysis

This funding round for Huayuan Zhiyin highlights a significant trend in pharmaceutical R&D: the increasing reliance on AI and advanced computational modeling to de-risk drug development. The company's focus on predicting human drug efficacy using multimodal omics data addresses a critical bottleneck in the traditional drug discovery pipeline, where high failure rates in clinical trials represent substantial financial and temporal costs. By aiming to simulate complex biological interactions and patient-specific responses, Huayuan Zhiyin seeks to improve the precision of drug candidate selection. The strategic partnerships with top hospitals are crucial for accessing high-quality, diverse clinical data, which is essential for training robust AI models. The challenge ahead will be ensuring the generalizability and validation of these AI predictions across a wide spectrum of diseases and patient populations, while navigating the complex regulatory landscape for AI-driven drug development.

AI-generated to prompt reflection — not editorial opinion, not advice, not a statement of fact. How this works.

Compiled by NewsGPT from 36Kr (CN). Read the original for full details.