北京：2025生物样本库建设技术攻关项目榜单 聚焦基因测序、蛋白质组学等

Core Viewpoint - The article discusses the announcement by the Beijing Municipal Science and Technology Commission and the Zhongguancun Science Park Management Committee regarding the 2025 technical tackling projects for the construction of biological sample libraries, focusing on three main areas: gene sequencing platforms, proteomics detection platforms, and artificial intelligence follow-up systems [2][23]. Gene Sequencing Platform - The goal is to establish a high-throughput gene sequencing platform that achieves full-process automation from sample to data [4]. - Key indicators include: - Automation level: complete automation of sample preprocessing and library construction systems [4]. - Throughput requirements: sequencing read length of at least 150 bp, compatible with targeted sequencing, whole genome sequencing, and whole genome methylation sequencing [5]. - Data quality: effective sequencing average depth of 30-35X for human whole genome sequencing, with average sequencing quality Q30 ≥ 85% and genome alignment rate ≥ 99% [6]. - Bioinformatics analysis capability: daily processing capacity of at least 120 human whole genome sequencing data [7]. - Cost control: exploring the limit cost of single sample gene sequencing based on 50,000 samples [8]. - Assessment indicators include establishing a complete gene sequencing system and providing data quality reports from third-party testing institutions, along with conducting gene testing services for no less than 5,000 samples [9]. Proteomics Detection Platform - The aim is to create a large-scale targeted proteomics detection platform with the following key indicators: - Automation level: full-process operation automation [10]. - Sample volume: capable of detecting samples as low as 5 µl [11]. - Throughput requirements: single detection throughput of no less than 300 samples per day, with an annual detection volume of no less than 70,000 [12]. - Concentration detection: achieving dynamic range detection of no less than five orders of magnitude [13]. - Data quality: capable of detecting at least 2,000 types of proteins in common clinical samples, with average CV ≤ 15% [14]. - Cost control: exploring the limit cost of single sample proteomics detection based on 50,000 samples [15]. - Assessment indicators include forming reagents and detection prototypes, providing testing reports to application units, and conducting proteomics detection services for no less than 1,000 samples [16]. Artificial Intelligence Follow-up System - The objective is to build a multi-channel follow-up system with the following key indicators: - Multi-channel follow-up: supporting intelligent outbound calls, SMS/WeChat, video, and wearable devices [17]. - High-fidelity multi-concurrent natural dialogue: voice recognition accuracy ≥ 97% and semantic understanding accuracy ≥ 97% [18]. - Support for smart wearable devices: integration of at least two types of wearable smart hardware with Class II medical device certification [19]. - System management capability: supporting management of at least 100 research cohorts, with a total patient capacity of no less than 100,000 [20]. - Patient services: precise push of relevant knowledge and support for closed-loop management of abnormal follow-ups [21]. - Assessment indicators include completing one set of the AI follow-up system, providing an operation manual and system testing report, and validating in real-world usage [22].