Search for: All records

Most neural networks assume that input images have a fixed number of channels (three for RGB images). However, there are many settings where the number of channels may vary, such as microscopy images where the number of channels changes depending on instruments and experimental goals. Yet, there has not been a systemic attempt to create and evaluate neural networks that are invariant to the number and type of channels. As a result, trained models remain specific to individual studies and are hardly reusable for other microscopy settings. In this paper, we present a benchmark for investigating channel-adaptive models in microscopy imaging, which consists of 1) a dataset of varied-channel single-cell images, and 2) a biologically relevant evaluation framework. In addition, we adapted several existing techniques to create channel-adaptive models and compared their performance on this benchmark to fixed-channel, baseline models. We find that channel-adaptive models can generalize better to out-of-domain tasks and can be computationally efficient. We contribute a curated dataset and an evaluation API to facilitate objective comparisons in future research and applications. 

Observing changes in cellular phenotypes under experimental interventions is a powerful approach for studying biology and has many applications, including treatment design. Unfortunately, not all interventions can be tested experimentally, which limits our ability to study complex phenomena such as combinatorial treatments or continuous time or dose responses. In this work, we explore unbiased, image-based generative models to analyze phenotypic changes in cell morphology and tissue organization. The proposed approach is based on generative adversarial networks (GAN) conditioned on feature representations obtained with self-supervised learning. Our goal is to ensure that image-based phenotypes are accurately encoded in a latent space that can be later manipulated and used for generating images of novel phenotypic variations. We present an evaluation of our approach for phenotype analysis in a drug screen and a cancer tissue dataset. 

Abstract Predicting assay results for compounds virtually using chemical structures and phenotypic profiles has the potential to reduce the time and resources of screens for drug discovery. Here, we evaluate the relative strength of three high-throughput data sources—chemical structures, imaging (Cell Painting), and gene-expression profiles (L1000)—to predict compound bioactivity using a historical collection of 16,170 compounds tested in 270 assays for a total of 585,439 readouts. All three data modalities can predict compound activity for 6–10% of assays, and in combination they predict 21% of assays with high accuracy, which is a 2 to 3 times higher success rate than using a single modality alone. In practice, the accuracy of predictors could be lower and still be useful, increasing the assays that can be predicted from 37% with chemical structures alone up to 64% when combined with phenotypic data. Our study shows that unbiased phenotypic profiling can be leveraged to enhance compound bioactivity prediction to accelerate the early stages of the drug-discovery process. 

Real-world deployment of computer vision systems, including in the discovery processes of biomedical research, requires causal representations that are invariant to contextual nuisances and generalize to new data. Leveraging the internal replicate structure of two novel single-cell fluorescent microscopy datasets, we propose generally applicable tests to assess the extent to which models learn causal representations across increasingly challenging levels of OODgeneralization. We show that despite seemingly strong performance as assessed by other established metrics, both naive and contemporary baselines designed to ward against confounding, collapse to random on these tests. We introduce a new method, Interventional Style Transfer (IST), that substantially improves OOD generalization by generating interventional training distributions in which spurious correlations between biological causes and nuisances are mitigated. We publish our code and datasets.