PO.ET02.06 · 实验与分子治疗
基于AI对数百万张IHC图像的分析识别出19对空间高度共表达的蛋白对,助力双特异性抗体开发
AI-powered analysis of millions of IHC images identifies 19 spatially highly co-expressed protein pairs to enable bispecific antibody development
作者与单位 Authors & Affiliations
摘要 Abstract
中文摘要
背景:开发有效的双特异性抗体需要识别在特定肿瘤相同空间背景中共表达的靶蛋白对。然而,由于跨越整个人类蛋白质组的组合搜索空间极为庞大,发现此类共表达靶点颇具挑战。基于AI对IHC图像的分析可实现对空间共表达蛋白对的系统性识别,从而促进合理的双特异性抗体设计。
方法:我们分析了来自Human Protein Atlas的680万张IHC染色的TMA核心图像,其中使用了21144种不同抗体对15303种人类蛋白进行染色。采用基于特征匹配的图像相似度算法识别连续切片对。对于每一对,使用Lunit SCOPE uIHC模型从IHC图像中检测阳性和阴性细胞。通过两种蛋白阳性细胞区域的交并比(IoU)量化其表达的空间相似度。
结果:在由150万张阳性染色的TMA核心中每两张图像组合生成的所有可能图像对中,基于图像相似度分析识别出26730对为连续切片。通过选择阳性细胞区域重叠(IoU ≥ 70)且仅由质膜蛋白组成的图像对,我们在10种肿瘤类型中识别出19对,其中膀胱癌中的配对最多。B7-H3在膀胱癌中与NT5E或JAG1共表达,在宫颈癌中与NT5E共表达。在甲状腺癌中,CLDN3与ROBO1共表达。
结论:我们开发了一套用于IHC图像空间分析的AI驱动流程,并将其应用于一个大型公开数据源,实现了对共表达蛋白对的系统性识别。该方法可指导肿瘤类型特异性双特异性抗体的设计,例如展示了同时靶向B7-H3和外核苷酸酶NT5E或Notch配体JAG1的可能性。
按肿瘤类型识别的共表达靶点对列表。肿瘤类型 配对数量 配对列表(A x B) 膀胱 7 B7-H3 x NT5E、B7-H3 x JAG1、PIEZO2 x GPR142、DDR1 x IL6ST、ITGA3 x BTN3A3、CSF1R x BTN3A3、LSR x SORT1 宫颈 2 B7-H3 x NT5E、DDR1 x ICOSLG 胰腺 2 SLC4A4 x PKD2L1、DCHS1 x GFRAL 脑 1 TRPV2 x CSPG4 头颈 1 DCHS1 x GFRAL 肝 1 CDH8 x KCNE3 肺 1 SLC4A4 x PKD2L1 睾丸 1 PIEZO2 x GPR142 甲状腺 1 CLDN3 x ROBO1 子宫 1 FZD1 x IGF1R
查看英文原文 English abstract
Background: Developing effective bispecific antibodies requires identifying target protein pairs that are co-expressed within the same spatial context of a given tumor. However, discovering such co-expressed targets is challenging due to the vast combinatorial search space across the human proteome. AI-powered analysis of IHC images enables systematic identification of spatially co-expressed protein pairs, facilitating rational bispecific antibody design.
Methods: We analyzed 6.8M IHC-stained TMA core images from the Human Protein Atlas, where 21144 different antibodies were used to stain 15303 human proteins. Serial section pairs were identified using a feature-matching-based image similarity algorithm. For each pair, positive and negative cells were detected from the IHC images using the Lunit SCOPE uIHC model. The spatial similarity of expression between two proteins was quantified using the intersection over union (IoU) of their positive cell regions.
Result: From all possible image pairs generated by combining every two images from 1.5M positively stained TMA cores, 26730 pairs were identified as serial sections based on image similarity analysis. By selecting pairs in which the positive cell regions overlapped (IoU ≥ 70) and only consisted of plasma membrane proteins, we identified 19 pairs across 10 tumor types with the most pairs in bladder cancer. B7-H3 was co-expressed with NT5E or JAG1 in bladder cancer, and with NT5E in cervical cancer. In thyroid cancer, CLDN3 was co-expressed with ROBO1.
Conclusion: We developed an AI-powered pipeline for spatial analysis of IHC images and applied it to a large publicly available data source, enabling systematic identification of co-expressed protein pairs. This approach can guide the design of tumor type specific bispecific antibodies as demonstrated by the possibility of targeting B7-H3 and simultaneously either the ectonucleotidase NT5E or the Notch ligand JAG1.
List of identified co-expressed target pairs by tumor types. Tumor type Number of pairs List of pairs (A x B) Bladder 7 B7-H3 x NT5E, B7-H3 x JAG1, PIEZO2 x GPR142, DDR1 x IL6ST, ITGA3 x BTN3A3, CSF1R x BTN3A3, LSR x SORT1 Cervix 2 B7-H3 x NT5E, DDR1 x ICOSLG Pancreas 2 SLC4A4 x PKD2L1, DCHS1 x GFRAL Brain 1 TRPV2 x CSPG4 Head and Neck 1 DCHS1 x GFRAL Liver 1 CDH8 x KCNE3 Lung 1 SLC4A4 x PKD2L1 Testis 1 PIEZO2 x GPR142 Thyroid 1 CLDN3 x ROBO1 Uterine 1 FZD1 x IGF1R
利益披露 Disclosure
S. Kim,
Lunit Inc. Employment, Stock Option.
H. Kim,
Lunit Inc. Employment.
B. Brattoli,
Lunit Inc. Employment, Stock, Stock Option.
S. Pereira,
Lunit Inc. Employment, Stock, Stock Option.
S. M. Ali,
Lunit Inc. Employment, Stock.
Revolution Medicines Inc. Stock.
IN8BIO Advisor.
Elevation Oncology Advisor.
Pillar Biosciences Advisor.
Promunity Advisor.
Droplet Biosciences Advisor.
Liquid Cell Advisor.
SRone Capital Management Advisor.