Accelerating Site-specific Characterization of Protein Therapeutics with Novel Machine Learning Methods

利用新型机器学习方法加速蛋白质治疗的位点特异性表征

• 批准号: 9927740
• 负责人: George Hall Johnson
• 金额: $ 48.53万
• 依托单位: MASSMATRIX, INC.
• 依托单位国家: 美国
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-05-06 至 2022-03-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/9927740
• 关键词: Accountability; Address; Biologic Development; Biological; Biological Sciences; Biomedical Research; Code; Communication; Communities; Complex Mixtures; Computer software; Data; Databases; Decision Making; Development; Directories; Disease; Drug Industry; Education; Ensure; Event; FOLH1 gene; Internet; Intervention; Life; Location; Machine Learning; Manuals; Mass Spectrum Analysis; Methods; Modification; Molecular; Monoclonal Antibodies; Mutation; Output; Pathway interactions; Peptide Mapping; Peptides; Phase; Process; Protein Databases; Proteins; Proteomics; Protocols documentation; Pythons; Readiness; Reliability of Results; Research; Resource Allocation; Resources; Sampling; Savings; Secure; Security; Shotguns; Site; Source; Speed; System; Systems Integration; Technology; Testing; Time; Variant; Visualization; base; biological systems; blind; candidate selection; commercialization; comparative; cost; data analysis pipeline; design; drug development; drug efficacy; encryption; experimental study; feature detection; improved; innovation; interoperability; lead candidate; light weight; machine learning method; medication safety; novel; prototype; quantum; repository; science education; therapeutic protein

PROJECT SUMMARY / ABSTRACT Title: Accelerating Site-specific Characterization of Protein Therapeutics with Novel Machine Learning Methods The project seeks to improve reliability and speed up the development of life-saving and life-enhancing, precision, protein therapeutics and magnify the positive impact of biomedical research and education worldwide, leading to a quantum leap in our understanding of the molecular and cellular pathways and mechanisms involved in healthy and diseased biological systems. The development of biologics is bottlenecked across the entire drug development process, from discovery to early stage candidate selection, process development and manufacturing, due to manual intervention in the mass spectrometry data analysis pipeline. Similarly, the proteomics research community is hindered as it moves from analysis of complex mixtures to more in-depth characterization of proteins and their modifications. Novel machine learning methods will be added to MassMatrix’s (LC-MS/MS software) proven analytical engine and visualization platform to minimize the loss of true positive peptide spectral matches. An innovative approach for efficient accountability of experimental data at the chromatogram level will also be researched, developed and added. The latter providing for easy traceability of each peak’s status in the chromatogram as soon as possible, thus providing convenient high-level assessment. Together, these aims are expected to improve the reliability and accuracy of results as well as to significantly reduce the mass spec bottleneck for the pharmaceutical industry and the research community. Deeper understanding and better decision making will follow, both having a potentially dramatic positive impact on downstream processes and resource deployment, including improved drug safety and efficacy.

项目摘要/摘要 标题：用新型机器学习方法加速蛋白质治疗的位点特异性表征 该项目旨在提高可靠性，加快救生和增强生命的发展， 精确，蛋白质治疗和放大生物医学研究和教育的积极影响 在世界范围内，导致我们对分子和细胞途径的理解发生了质的飞跃， 健康和疾病生物系统中的机制。生物制品的发展是 在整个药物开发过程中，从发现到早期候选药物的选择， 由于质谱数据分析中的人工干预，工艺开发和生产 渠道.类似地，蛋白质组学研究团体在从复杂的蛋白质组学分析 混合物，以更深入地表征蛋白质及其修饰。 新型机器学习方法将添加到MassMatrix（LC-MS/MS软件）经过验证的分析引擎中 和可视化平台，以最大限度地减少真阳性肽谱匹配的损失。一个创新 还将研究在色谱水平上有效计算实验数据的方法， 开发和添加。后者提供了色谱图中每个峰状态的简单可追溯性， 尽快，从而提供方便的高水平评估。这些目标预计将共同 提高结果的可靠性和准确性，并显著降低质谱瓶颈， 制药业和研究界。更深入的理解和更好的决策 这两个方面都可能对下游工艺和资源产生巨大的积极影响。 部署，包括提高药物安全性和有效性。