Wang, R.-S. | bioRxiv Science

bioRxiv2024

StomaVision: stomatal trait analysis through deep learning

StomaVision is an automated tool designed for high-throughput detection and measurement of stomatal traits, such as stomatal number, pore size, and closure rate. It provides insights into plant responses to environmental cues, streamlining the analysis of micrographs from field-grown plants across various species, including monocots and dicots. Enhanced by a novel collection method that utilizes video recording, StomaVision increases the number of captured images for robust statistical analysis. Accessible via an intuitive web interface at and available for local use in a containerized environment at , this tool ensures long-term usability by minimizing the impact of software updates and maintaining functionality with minimal setup requirements. The application of StomaVision has provided significant physiological insights, such as variations in stomatal density, opening rates, and total pore area under heat stress. These traits correlate with critical physiological processes, including gas exchange, carbon assimilation, and water use efficiency, demonstrating the tools utility in advancing our understanding of plant physiology. The ability of StomaVision to identify differences in responses to varying durations of heat treatment highlights its value in plant science research. Plain language summaryStomaVision is a tool that automatically counts and measures tiny openings on plant leaves, helping us learn how plants deal with their surroundings. It is easy to use and works well with various plant species. This tool helps scientists see how plants change under stress, making plant research easier and more accurate.

plant biology↗

bioRxiv2018

Single-Cell Modeling of CD8+ T Cell Exhaustion Predicts Response to Cancer Immunotherapy

Accurate prediction of response to immune checkpoint blockade (ICB) and simultaneous development of new efficacious ICB strategy are unmet needs for cancer immunotherapy1-5. Despite the advance of existing clinical or computational biomarkers6-10, it is very limited in explicating phenotypic variability of tumor-infiltrating CD8+ T lymphocytes (TILs) at the single-cell level, including tumor-specificity11,12 and distinct exhaustion profiles13-16, that are crucial for the responsiveness to ICB. Here we show that a quantitative criterion, D value, for evaluating tumor-specific TIL exhaustion, identified by our new high-dimensional single-cell-based computational method, called HD-scMed, accurately predicts response to the ICB, i.e., PD-1, CTLA-4, and PD-1+CTLA-4, with the performance of AUC=100% in human tumors. D-value is the Euclidean distance from a subset of exhausted CD8+ TILs, identified as Pareto Front (PF)18-21 by HD-scMed within the high-dimensional expression space of a variety of exhaustion markers, to the baseline TILs excluded by the PF. We phenotypically distinguished two types of TIL \"exhaustion\" by the D value, namely \"D Extremely-high\" in non-responders, associating with high tumor-specificity imprinted with enhanced exhaustion and inactivated effector and cytotoxic signatures; and \"D low\" specific to responders, alternatively enriched with T cell activation and cytolytic effector T cell signatures. We also observed a large portion of \"D negative\" bystander T cells irrelevant to response. Notably, D-low TILs in clinical responders display very low LAG3 expression. To reverse the functionality of the LAG3-high TILs after receiving PD-1+CTLA-4, we combined LAG3 with PD-1+CTLA-4 in treating a murine tumor model. Remarkably, LAG3+PD-1+CTLA-4 displays extraordinary antitumor efficacy to eradicate advanced tumors, which is associated with burst GZMBhi TIL populations; whereas PD-1+CTLA-4 or PD-1+CTLA-4 plus other ICBs as control only induce moderate tumor growth inhibition. Our study has important implications for cancer immunotherapy as providing both accurate predictions of ICB response and alternative strategy to reverse ICB resistance.

bioinformatics↗