Variational and Explanatory Neural Networks for Encoding Cancer Profiles and Predicting Drug Responses
George Mason University(乔治梅森大学);Mayo Clinic(梅奥诊所);Argonne National Laboratory(阿贡国家实验室)
Tianshu Feng, Rohan Gnanaolivu, Abolfazl Safikhani, Yuanhang Liu, Jun Jiang, Nicholas Chia, Alexander Partin, Priyanka Vasanthakumari, Yitan Zhu, Chen Wang
本文提出 VETE 神经网络框架,通过变分组件降噪并将基因本体可追溯地融入架构,提升癌细胞系分类和药物反应预测的准确性与生物学可解释性。
摘要:
人类癌症构成了重大的公共卫生挑战,需要通过转化研究发现新药。描述肿瘤和癌细胞系中分子活动的转录组学谱数据被广泛用于预测抗癌药物反应。然而,现有的人工智能模型面临转录组学数据中的噪声和缺乏生物学可解释性等挑战。为克服这些局限,我们提出了 VETE(Variational and Explanatory Transcriptomics Encoder,变分与可解释转录组学编码器),这是一种新型神经网络框架,它引入变分组件以减轻噪声影响,并将可追溯的基因本体整合到神经网络架构中,用于编码癌症转录组学数据。关键创新包括:一种用于识别本体路径的局部可解释性引导方法、一种阐明药物反应生物学机制的可视化工具,以及集中式大规模超参数优化的应用。VETE 在癌细胞系分类和药物反应预测中表现出稳健的准确性。此外,它为这两项任务提供了可追溯的生物学解释,并为其预测背后的机制提供了见解。VETE 弥合了人工智能驱动的预测与癌症研究中具有生物学意义的见解之间的差距,代表了该领域一项有前景的进展。
英文摘要:
Human cancers present a significant public health challenge and require the discovery of novel drugs through translational research. Transcriptomics profiling data that describes molecular activities in tumors and cancer cell lines are widely utilized for predicting anti-cancer drug responses. However, existing AI models face challenges due to noise in transcriptomics data and lack of biological interpretability. To overcome these limitations, we introduce VETE (Variational and Explanatory Transcriptomics Encoder), a novel neural network framework that incorporates a variational component to mitigate noise effects and integrates traceable gene ontology into the neural network architecture for encoding cancer transcriptomics data. Key innovations include a local interpretability-guided method for identifying ontology paths, a visualization tool to elucidate biological mechanisms of drug responses, and the application of centralized large scale hyperparameter optimization. VETE demonstrated robust accuracy in cancer cell line classification and drug response prediction. Additionally, it provided traceable biological explanations for both tasks and offers insights into the mechanisms underlying its predictions. VETE bridges the gap between AI-driven predictions and biologically meaningful insights in cancer research, which represents a promising advancement in the field.
Enhancing drug and cell line representations via contrastive learning for improved anti-cancer drug prioritization
The Ohio State University(俄亥俄州立大学);Biomedical Informatics Department, The Ohio State University(俄亥俄州立大学生物医学信息学系);Computer Science and Engineering Department, The Ohio State University(俄亥俄州立大学计算机科学与工程系);Translational Data Analytics Institute, The Ohio State University(俄亥俄州立大学转化数据分析研究所)
TransCDR: a deep learning model for enhancing the generalizability of cancer drug response prediction through transfer learning and multimodal data fusion for drug representation
Shanghai Medical College, Fudan University(复旦大学上海医学院);Shanghai Institute of Stem Cell Research and Clinical Translation(上海干细胞研究与临床转化研究院);Shanghai World Foreign Language Academy(上海世界外国语中学)
TCR: A Transformer Based Deep Network for Predicting Cancer Drugs Response
Baidu(百度)
由于肿瘤的异质性,在个性化的基础上预测抗癌药物的临床结果在癌症治疗中具有挑战性。传统的计算工作是通过分子谱来描述药物对个体样本的影响,但由于组学数据的高维性,会出现过拟合,阻碍了模型的临床应用。最近的研究表明,深度学习是一种很有前途的方法,通过学习药物和样本之间的比对模式来建立药物反应模型。然而,现有的研究采用简单的特征融合策略,只考虑作为一个整体表示的药物特征,而忽略了子结构信息,可能发挥重要作用时,比对药物和基因。因此,在本文中,我们提出TCR(Transformer based network for Cancer drug Response)来预测抗癌药物的反应。通过利用注意力机制,TCR能够有效地学习药物原子/亚结构与分子特征之间的相互作用。此外,设计了双损失函数和交叉采样策略,以提高TCR的预测能力。我们表明,TCR优于所有其他方法下的各种数据分割策略的所有评估矩阵(一些显着改善)。大量的实验表明,TCR在独立的体外实验和体内真实患者数据上显示出显着提高的泛化能力。我们的研究强调了TCR的预测能力及其对癌症药物再利用和精确肿瘤治疗的潜在价值。
A Drug Recommendation System (Dr.S) for cancer cell lines
Centrum Wiskunde & Informatica(数学与计算机科学中心);Utrecht University(乌得勒支大学);Massachusetts General Hospital(麻省总医院);Harvard Medical School(哈佛医学院);Maastricht University Medical Centre+(马斯特里赫特大学医学中心+)
Ensemble Transfer Learning for the Prediction of Anti-Cancer Drug Response
Argonne National Laboratory(阿贡国家实验室);Leidos Biomedical Research, Inc.(莱多斯生物医学研究公司);National Cancer Institute(美国国家癌症研究所);The University of Chicago(芝加哥大学)
Evaluating the Effectiveness of Artificial Intelligence in Predicting Adverse Drug Reactions among Cancer Patients: A Systematic Review and Meta-Analysis
National Polytechnic School of Oran - Maurice Audin(奥兰国家理工学院 - 莫里斯·奥丹)
Permutation invariant multi-output Gaussian Processes for drug combination prediction in cancer
MRC Biostatistics Unit, University of Cambridge(剑桥大学医学研究委员会生物统计学单元);Oslo Centre for Biostatistics and Epidemiology (OCBE), University of Oslo(奥斯陆大学奥斯陆生物统计学与流行病学中心);Broad Institute of MIT & Harvard(麻省理工学院与哈佛大学博德研究所);CRUK Cambridge Centre Ovarian Programme(英国癌症研究剑桥中心卵巢项目);Cambridge Institute of Therapeutic Immunology and Infectious Disease (CITIID)(剑桥治疗免疫学与传染病研究所)
Learning Curves for Drug Response Prediction in Cancer Cell Lines
Argonne National Laboratory(阿贡国家实验室);University of Chicago(芝加哥大学);Frederick National Laboratory for Cancer Research(弗雷德里克国家癌症研究实验室);Leidos Biomedical Research, Inc.(Leidos生物医学研究公司)
VQSynery: Robust Drug Synergy Prediction With Vector Quantization Mechanism
School of Data Science and Engineering(数据科学与工程学院);East China Normal University(华东师范大学);School of Medicine(医学院);College of Design&Engineering(设计与工程学院);National University of Singapore(新加坡国立大学)
Pathway-Guided Optimization of Deep Generative Molecular Design Models for Cancer Therapy
Texas A&M University(德克萨斯农工大学);Frederick National Laboratory for Cancer Research(弗雷德里克国家癌症研究所);Leidos Biomedical Research(莱多斯生物医学研究所);Brookhaven National Laboratory(布鲁克海文国家实验室)
Adaptive High-Level Tight Control of Prostate Cancer: A Path from From Terminal Disease to Chronic Condition
Scripps and Pitzer Colleges, Claremont Colleges Consortium(斯克里普斯学院和皮策学院,克莱蒙特学院联盟);Princeton University(普林斯顿大学);Scripps College(斯克里普斯学院);Cancer Ecology Center, The Brady Urological Institute, Johns Hopkins School of Medicine(约翰霍普金斯医学院布雷泌尿研究所癌症生态学中心);Moffitt Cancer Center(莫菲特癌症中心);Johns Hopkins University(约翰霍普金斯大学);Bloomberg School of Public Health, Johns Hopkins University(约翰霍普金斯大学布隆伯格公共卫生学院);Department of Chemical and Biomolecular Engineering, John Hopkins University(约翰霍普金斯大学生物化学工程系)