Search

A wide variety of diseases are commonly diagnosed via the visual examination of cell morphology within a peripheral blood smear. For certain diseases, such as COVID-19, morphological impact across the multitude of blood cell types is still poorly understood. In this paper, we present a multiple instance learning-based approach to aggregate high-resolution morphological information across many blood cells and cell types to automatically diagnose disease at a per-patient level. We integrated image and diagnostic information from across 236 patients to demonstrate not only that there is a significant link between blood and a patient’s COVID-19 infection status, but also that novel machine learning approaches offer a powerful and scalable means to analyze peripheral blood smears. Our results both backup and enhance hematological findings relating blood cell morphology to COVID-19, and offer a high diagnostic efficacy; with a 79% accuracy and a ROC-AUC of 0.90. Author summary In this paper, we present a multiple instance learning-based approach to aggregate high-resolution morphological information across many blood cells and cell types to automatically diagnose COVID-19 at a per-patient level. We integrated image and diagnostic information from 236 patients to demonstrate not only that there is a significant link between blood and a patient’s COVID-19 infection status, but also that novel machine learning approaches offer a powerful and scalable means to analyze peripheral blood smears. Our results both backup and enhance hematological findings relating blood cell morphology to COVID-19, and offer high diagnostic accuracy. Besides the final aggregated decision, the proposed attention mechanism also provides cell-type importance, which can help pathologists to build valuable insights on which cell types are more diagnostically relevant, opening a window into improving the explainability of deep optical blood analysis approaches.

The human DDB1-CUL4 ASSOCIATED FACTOR (DCAF) proteins have been reported to interact directly with UV-DAMAGED DNA BINDING PROTEIN1 (DDB1) through the WDxR motif in their WD40 domain and function as substrate-recognition receptors for CULLIN4-based E3 ubiquitin ligases. Here, we identified and characterized a homolog of human DCAF1/VprBP in Arabidopsis thaliana. Yeast two-hybrid analysis demonstrated the physical interaction between DCAF1 and DDB1 from Arabidopsis, which is likely mediated via the WD40 domain of DCAF1 that contains two WDxR motifs. Moreover, coimmunoprecipitation assays showed that DCAF1 associates with DDB1, RELATED TO UBIQUITIN-modified CUL4, and the COP9 signalosome in vivo but not with CULLIN-ASSOCIATED and NEDDYLATION-DISSOCIATED1, CONSTITUTIVE PHOTOMORPHOGENIC1 (COP1), or the COP10-DET1-DDB1 complex, supporting the existence of a distinct Arabidopsis CUL4 E3 ubiquitin ligase, the CUL4-DDB1-DCAF1 complex. Transient expression of fluorescently tagged DCAF1, DDB1, and CUL4 in onion epidermal cells showed their colocalization in the nucleus, consistent with the notion that the CUL4-DDB1-DCAF1 complex functions as a nuclear E3 ubiquitin ligase. Genetic and phenotypic analysis of two T-DNA insertion mutants of DCAF1 showed that embryonic development of the dcaf1 homozygote is arrested at the globular stage, indicating that DCAF1 is essential for plant embryogenesis. Reducing the levels of DCAF1 leads to diverse developmental defects, implying that DCAF1 might be involved in multiple developmental pathways. [ABSTRACT FROM AUTHOR]