Open AccessTrans-channel fluorescence learning improves high-content screening for Alzheimer’s disease therapeutics
Author(s) -
Daniel Wong,
Jay C. Conrad,
Noah R. Johnson,
Jacob I. Ayers,
Annelies Laeremans,
Joanne C. Lee,
Sang Hoon Lee,
Stanley B. Prusiner,
Sourav Bandyopadhyay,
Atul J. Butte,
Nick A. Paras,
Michael J. Keiser
Publication year - 2022
Publication title -
nature machine intelligence
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 4.894
H-Index - 16
ISSN - 2522-5839
DOI - 10.1038/s42256-022-00490-8
Subject(s) - generalizability theory , pipeline (software) , computer science , high content screening , fluorescence , high throughput screening , drug discovery , artificial intelligence , machine learning , computational biology , bioinformatics , chemistry , biology , biochemistry , cell , mathematics , statistics , physics , quantum mechanics , programming language
In microscopy-based drug screens, fluorescent markers carry critical information on how compounds affect different biological processes. However, practical considerations, such as the labor and preparation formats needed to produce different image channels, hinders the use of certain fluorescent markers. Consequently, completed screens may lack biologically informative but experimentally impractical markers. Here, we present a deep learning method for overcoming these limitations. We accurately generated predicted fluorescent signals from other related markers and validated this new machine learning (ML) method on two biologically distinct datasets. We used the ML method to improve the selection of biologically active compounds for Alzheimer's disease (AD) from a completed high-content high-throughput screen (HCS) that had only contained the original markers. The ML method identified novel compounds that effectively blocked tau aggregation, which had been missed by traditional screening approaches unguided by ML. The method improved triaging efficiency of compound rankings over conventional rankings by raw image channels. We reproduced this ML pipeline on a biologically independent cancer-based dataset, demonstrating its generalizability. The approach is disease-agnostic and applicable across diverse fluorescence microscopy datasets.