生物医药研发
Search documents
《科学》杂志发表中国科研新成果 为精准医学提供新工具
Xin Lang Cai Jing· 2026-01-04 03:30
来源:人民日报海外版 据新华社杭州1月2日电(记者黄筱)我国科学家成功构建了能够在单细胞分辨率下,同步实现细胞膜表 面标志物发现与靶向核酸适体探针获取的一体化平台——SPARK-seq。该技术为分子识别、原创靶点发 现及精准医学研究提供了全新工具,有望为众多因靶点不明而缺乏有效疗法的疾病开辟新的治疗途径。 由中国科学院杭州医学研究所谭蔚泓院士与吴芩研究员团队合作取得的这项研究成果,1月1日在国际学 术期刊《科学》上线。 细胞膜蛋白是药物作用的关键靶点,而核酸适体是一类能够高特异性、高亲和力结合靶标分子的寡核苷 酸。然而,传统的核酸适体筛选方法效率低、过程繁杂,且难以在生理相关环境下系统发现全新的疾病 标志物。 SPARK-seq平台的诞生解决了这一挑战。相较于传统方法,其筛选效率提升达百倍以上,并能高精度锁 定潜在的癌症标志物与治疗靶点。吴芩表示,"该平台在单细胞层面部署的'分子雷达',能够大规模、 并行地精准识别细胞表面与疾病相关的靶标,并同步获取能特异性结合它的核酸适体探针。" SPARK-seq作为自主可控的原创性平台,为"无药可靶"疾病的靶点研究与治疗开发提供了全新范式,使 得科研与临床工作者能够基于 ...
张强锋/汪阳明合作开发AI工具SMRTnet,无需RNA三级结构,精准预测小分子-RNA相互作用
生物世界· 2026-01-03 09:30
Core Viewpoint - RNA-targeted small molecule drugs are emerging as a new frontier in the biopharmaceutical field, addressing the challenge of lacking precise tertiary structure information for most disease-related RNAs, which traditional computational methods rely on for predicting interactions with small molecules [2][5]. Group 1: SMRTnet Development - The research teams from Tsinghua University and Peking University developed a deep learning tool named SMRTnet, which predicts small molecule-RNA interactions without relying on RNA tertiary structures, thus expanding the range of targetable RNA [2][5]. - SMRTnet integrates two large language models, convolutional neural networks, and graph attention networks, utilizing only RNA sequence and secondary structure information to predict binding capabilities [5]. Group 2: Performance and Validation - SMRTnet demonstrated robust performance, achieving an average area under the receiver operating characteristic curve (auROC) of 0.830-0.844 in five-fold cross-validation, significantly outperforming existing tools [9]. - In bait evaluation tasks, SMRTnet's average ranking reached 92.6%, far exceeding four molecular docking tools (27.3%-46.6%) and two deep learning tools (16.0%-23.8%), indicating its superior ability to identify true binding molecules [9]. - The model's performance was notably affected when using predicted secondary structures instead of experimentally determined ones, highlighting the importance of accurate experimental data [10]. Group 3: Binding Site Prediction - SMRTnet can also identify small molecule binding sites on RNA, quantifying the contribution of each nucleotide to the predicted binding score using the Grad-CAM algorithm [12]. - The accuracy of binding site predictions was validated against experimentally determined sites, achieving an average auROC of 0.695-0.793 across multiple datasets [13]. - Out of 190 predicted small molecule-RNA interactions, 40 were experimentally validated as binding molecules, yielding an average validation rate of 21.1% [14]. Group 4: Case Study on MYC IRES - The research focused on MYC IRES, a target considered "undruggable," showing a positive correlation between predicted binding scores and experimental validation rates, with a validation rate of 28.6% for scores between 0.9-1.0 [16]. - Among 15 identified MYC IRES binders, the team highlighted IHT (Irinotecan Hydrochloride Trihydrate) for its favorable drug development characteristics [17]. - IHT was predicted to bind at a specific site on MYC IRES, and experimental validation confirmed the reliability of SMRTnet's predictions, demonstrating significant reductions in MYC mRNA and protein levels in HeLa cells [19]. Group 5: Future Prospects - SMRTnet represents a significant advancement in predicting small molecule-RNA interactions, overcoming traditional method limitations and broadening the scope of disease-related RNA targets [21]. - The accumulation of multi-omics data, including chemical-RNA interaction genomics and functional screening, is expected to enhance AI methods for predicting binding interactions and downstream biological effects, accelerating RNA-targeted drug development [21]. - The study illustrates the potential of AI-driven approaches in RNA-targeted small molecule therapy, with expectations for breakthrough advancements in the coming years as data quality and algorithm optimization improve [21].
《科学》杂志发表我国科研新成果 为精准医学提供新工具
Xin Hua She· 2026-01-02 09:07
研究团队在实验室讨论。(资料图) SPARK-seq作为自主可控的原创性平台,为"无药可靶"疾病的靶点研究与治疗开发提供了全新范式,使 得科研与临床工作者能够基于我国独特的临床资源,系统性地发现具有自主知识产权的新靶点与新工 具。未来,该技术有望推动形成一种"按图索骥"式的精准医疗新模式,快速为不同疾病匹配或定制治疗 分子,提升药物研发的效率和治疗方案的精准度。 细胞膜蛋白是药物作用的关键靶点,而核酸适体是一类能够高特异性、高亲和力结合靶标分子的寡核苷 酸。然而,传统的核酸适体筛选方法效率低、过程繁杂,且难以在生理相关环境下系统发现全新的疾病 标志物。 SPARK-seq平台的诞生解决了这一挑战。相较于传统方法,其筛选效率提升达百倍以上,并能高精度锁 定潜在的癌症标志物与治疗靶点。"我们研发SPARK-seq的初衷,是为了攻克像三阴性乳腺癌这类缺乏 明确治疗靶点的临床难题。"吴芩表示,"该平台在单细胞层面部署的'分子雷达',能够大规模、并行地 精准识别细胞表面与疾病相关的靶标,并同步获取能特异性结合它的核酸适体探针。" 我国科学家成功构建了能够在单细胞分辨率下,同步实现细胞膜表面标志物发现与靶向核酸适体探针获 ...
中欧班列(武汉)开行“新年班列”,直抵丹麦首都哥本哈根
Chang Jiang Ri Bao· 2026-01-02 01:04
武汉多次组织企业代表团赴哥本哈根开展经贸交流活动,与丹麦工业联合会、哥本哈根投资促进署等机构建立联系机制。同时,哥本 哈根的企业和机构,也积极参与在武汉举办的中国—北欧经贸合作论坛,推动双方在绿色转型、智能制造、消费品等领域合作。 中欧班列(武汉)开行新线路直达哥本哈根线路图 2026年1月1日,迎着新年的第一缕曙光,一列满载新风换气机等货物的中欧班列(武汉),从中铁联集武汉中心站鸣笛启程,前往丹 麦首都哥本哈根。 这是湖北港口汉欧国际在新年开通的首条"新年班列",途经阿拉山口口岸出境,抵达德国汉堡,部分货物将依托成熟的"铁水联运"模 式转运分拨,最终抵达哥本哈根,全程运行时效预计22天。 此前,班列已开通芬兰赫尔辛基,挪威奥斯陆、莫斯,瑞典哥德堡等多条线路;如今直达哥本哈根,不仅标志着通道网络向北欧五国 深度延伸,进一步完善了湖北与北欧的跨境物流体系。 新年首趟中欧班列(武汉) 哥本哈根,意为"商人港口",是北欧最大城市,拥有造船、机器制造、冶金、化学、食品加工和纺织等工业,向外输出肉类和奶制 品。 武汉与丹麦之间贸易规模持续增长,参与了大量对丹出口业务,出口产品包括运输设备、电信产品、机械等,进口则以乳制 ...
科学圆桌会·趣谈2025| 药理学家:这一年,国产创新药正在经历“DeepSeek时刻”
Xin Hua She· 2025-12-31 05:04
今年,我们团队经过多年的努力,提出了靶向肾脏纤维化的嵌合抗原受体T细胞免疫疗法(CAR-T)新 思路,引起了业界的高度关注。但我深知,这仅仅是中国药物研发与细胞治疗领域快速发展大潮中的一 朵小小的浪花。 有一天,我们团队与国内生物医药公司讨论完这一新疗法的临床研究方案后,已是午夜时分。走出实验 室,一直紧绷的神经放松下来,我才注意到冬夜的校园那么美,多年前栽下的蜡梅已含苞待放。这何尝 不是创新药从零起步、艰难"绽放"的写照? 身为医药人,站在2025年岁末,有一种格外强烈的感慨:从被业界誉为"中国创新药元年"的2015年算 起,十年磨一剑,国产创新药正在经历"DeepSeek时刻":以长期积累的创新努力迎来产品重大突破。 在医药界,创新药有两个定律。 一个是"双十定律":十年时间、十亿美元,才能让一个新药从实验室 走向患者。这道"高墙",曾让无数创新梦想折戟;另一个就是"九死一生定律":约90%的创新药项目在 临床前或临床阶段失败,最终仅少数获批上市。 就拿我研究的领域来说,慢性肾脏病(CKD)正成为全球公共卫生面临的新挑战。今年5月召开的第78 届世界卫生大会(WHA),将肾脏疾病列入全球优先关注的重大非传 ...
药理学家:这一年,国产创新药正在经历“DeepSeek时刻”
Xin Hua She· 2025-12-31 05:02
身为医药人,站在2025年岁末,有一种格外强烈的感慨:从被业界誉为"中国创新药元年"的2015年算 起,十年磨一剑,国产创新药正在经历"DeepSeek时刻":以长期积累的创新努力迎来产品重大突破。 今年,我们团队经过多年的努力,提出了靶向肾脏纤维化的嵌合抗原受体T细胞免疫疗法(CAR-T)新 思路,引起了业界的高度关注。但我深知,这仅仅是中国药物研发与细胞治疗领域快速发展大潮中的一 朵小小的浪花。 有一天,我们团队与国内生物医药公司讨论完这一新疗法的临床研究方案后,已是午夜时分。走出实验 室,一直紧绷的神经放松下来,我才注意到冬夜的校园那么美,多年前栽下的蜡梅已含苞待放。这何尝 不是创新药从零起步、艰难"绽放"的写照? 在医药界,创新药有两个定律。 一个是"双十定律":十年时间、十亿美元,才能让一个新药从实验室 走向患者。这道"高墙",曾让无数创新梦想折戟;另一个就是"九死一生定律":约90%的创新药项目在 临床前或临床阶段失败,最终仅少数获批上市。 就拿我研究的领域来说,慢性肾脏病(CKD)正成为全球公共卫生面临的新挑战。今年5月召开的第78 届世界卫生大会(WHA),将肾脏疾病列入全球优先关注的重大非传 ...
英矽智能(3696)香港公開發售超購1427倍 基石投資者認購佔比近四成 擬於12月30日上市
Xin Lang Cai Jing· 2025-12-29 16:28
来源:新浪港股-好仓工作室 本次全球發售的聯席保薦人為摩根士丹利、中國國際金融香港證券有限公司及廣發證券(香港)金融控 股有限公司。英矽智能的股份預計將於2025年12月30日(星期二)上午九時正開始在聯交所買賣,股份 代號為3696,每手買賣單位為500股。 点击查看公告原文>> 声明:市场有风险,投资需谨慎。 本文为AI大模型基于第三方数据库自动发布,任何在本文出现的信 息(包括但不限于个股、评论、预测、图表、指标、理论、任何形式的表述等)均只作为参考,不构成 个人投资建议。受限于第三方数据库质量等问题,我们无法对数据的真实性及完整性进行分辨或核验, 因此本文内容可能出现不准确、不完整、误导性的内容或信息,具体以公司公告为准。如有疑问,请联 系biz@staff.sina.com.cn。 英矽智能公布全球發售結果,最終發行價定為每股24.05港元。本次全球發售的發售股份數目為 94,690,500股,視乎超額配股權行使與否而定,其中香港發售股份數目為9,469,500股,國際發售股份數 目為85,221,000股。按發行價計算,所得款項總額約為22.77億港元,扣除估計應付上市開支2.52億港元 後,所得 ...
新股暗盘 | 英矽智能(03696)暗盘收涨50.1% 每手赚6025港元
智通财经网· 2025-12-29 10:38
| 18:34 | | | | .Il ? D | | --- | --- | --- | --- | --- | | 首页 | | 英矽智能 HK 03696 暗盘 | | Q C | | 行情 | 板块 | 评论 | 券商持仓 | 实时数 | | 36.100 | | 成交 356.00万股 最高 72.300 | | 今开 48.960 | | +50.10% +12.050 换手 0.64% | | | 最低 31.000 | 总市 201.23亿 | 分时量 量:0 25.80万 36.020 36.000 450( 35.980 35.960 MACD MACD(12,26,9):0.009 DIF :- 0.058 DEA:- 0.063 35.940 35.920 35.900 35.880 16:15 18:30 分时 1分 5分 日K 立即买入 立即卖出 • 行情来源: 利弗莫尔证券 • 智通财经APP获悉,英矽智能(03696)将于2025年12月30日(星期二)在香港挂牌。截至收盘,利弗莫尔证券暗盘交易显示报价36.1港元,较招股价24.05港元上 涨50.1%,每手500股,不计手续 ...
新股暗盘 | 英矽智能(03696)暗盘盘初涨逾62% 每手赚7475港元
智通财经网· 2025-12-29 08:58
ZU.09/J 量 : 13000股 额 : 50.68万 16:15 17:00 17:30 18:00 18:30 分时 5分 1分 日K • 行情来源: 利弗莫尔证券 • | 英矽智能 | | | | | --- | --- | --- | --- | | ರ ೧ | ( | HK 03696 暗盘 | | | 39.000 | | | | | 最高 72.300 | 今开 48.960 | 成交量 112.0万股 | | | 总市 217.4亿 | 最低 38.500 | 换手 0.20% | +62.16% +14.950 | | 20251229 16:23 价 39.000 62.16% 均 43.017 分时量 13000股 | | | | | 52.301 | 117.47% | 十档 | 成交明细 | | 39.200 | - - - (- -) | | | | 39.180 | - - - - - - - ) | | | | - -(- -) | 39.160 | | | | - | 39.140 | | | | - | 39.120 | | | | 2000(4家) | 39.100 ...
多家巨头从美国私有化退市,中概股加速回归!
证券时报· 2025-12-28 12:59
Core Viewpoint - The Chinese concept stock market is undergoing significant changes in 2025, characterized by a wave of privatizations and delistings from U.S. exchanges, while a number of small and medium-sized enterprises continue to seek global financing opportunities, particularly through listings in the U.S. and Hong Kong [3][4]. Group 1: Privatization and Delisting - Geely Automobile completed the privatization of Zeekr, which became a wholly-owned subsidiary and delisted from the NYSE. The privatization was marked by a rapid process, with 70.8% of Zeekr shareholders opting for shares and 29.2% for cash, totaling $701 million [6]. - Dada Group, part of the JD ecosystem, was privatized by JD Group at a valuation of $520 million, allowing for more strategic flexibility and deeper collaboration with JD in the instant retail market [7]. - Financial One Account pioneered dual delisting by completing its exit from both the NYSE and Hong Kong Stock Exchange, with a privatization deal valued at approximately HKD 1.69 billion, driven by long-term low stock prices and liquidity issues [7]. Group 2: Trends in U.S. Listings - In 2025, 63 Chinese companies went public in the U.S., raising approximately $1.12 billion, indicating a trend of increasing numbers but decreasing fundraising amounts, with an average fundraising of less than $20 million [9]. - The largest IPOs included Bawang Tea and Ascentage Pharma, raising $411 million and $126 million respectively, highlighting a shift towards smaller enterprises in the U.S. market [9][10]. - The outlook for 2026 is cautious, as new listing requirements from Nasdaq may lead to a decline in the number of Chinese companies able to meet these standards [10]. Group 3: Return to Hong Kong - The trend of Chinese companies returning to Hong Kong is gaining momentum, with companies like Pony.ai and Hesai achieving dual primary listings, which is becoming the mainstream return model [12]. - Hesai's IPO in Hong Kong was the largest in the global lidar industry to date, raising over HKD 4.16 billion (approximately $533 million) [12]. - Other companies, such as Tianjing Biopharma, are also planning to pursue dual listings in Hong Kong, indicating a broader trend of returning to Asian markets [12]. Group 4: Strategic Implications - Some analysts suggest that privatization followed by IPOs in Hong Kong or A-shares may allow companies to escape U.S. regulatory pressures and achieve better valuations in local markets [13].