Machine Learning

• NYSCF has built a deep learning framework for automatically detecting clonalized cell colonies and assessing clonality from daily imaging data. Our approach, Monoqlo℠, features multiple CNN “modules” for both detection and classification, which are integrated and automatically deployed using Python scripts. Full details on the design and rationale of the Monoqlo℠ framework are provided in this article.

• NYSCF, in collaboration with Google Accelerated Science has built a high-throughput, high-content Cell Painting–based phenotyping platform that combines advanced and scalable cell culture automation with cutting-edge deep learning algorithms. This platform uses deep learning, in parallel with automated Cell Painting analysis, to confidently separate fibroblasts from subjects with Parkinson’s disease (PD) (both sporadic and LRRK2, ROC AUC 0.79 (0.08 standard deviation (SD)) from demographically matched healthy controls, demonstrating the potential use for this platform for unbiased PD disease modeling and drug discovery. Furthermore, this platform is able to identify a cell line within a cohort of 96 total lines with 91% mean accuracy (6% SD), across batches and plate layouts, demonstrating the robustness of the screening platform. Full details of our platform and study are summarized in this article.

Data Sets

• Monoqlo: We provide a demonstrative example of an imaging dataset, which includes an illustration of the Monoqlo℠ framework’s execution logic and performance. Visit the Monoqlo page for more information and to download the sample test dataset.
  
• AD-PD: We provide a Fiji-based macro to assess the quality and consistency of images acquired from a 96-well plate. Visit the ADPD page for more information and to download a sample test dataset.