Search

Artificial intelligence (AI) has shown promise for diagnosing prostate cancer in biopsies. However, results have been limited to individual studies, lacking validation in multinational settings. Competitions have been shown to be accelerators for medical imaging innovations, but their impact is hindered by lack of reproducibility and independent validation. With this in mind, we organized the PANDA challenge-the largest histopathology competition to date, joined by 1,290 developers-to catalyze development of reproducible AI algorithms for Gleason grading using 10,616 digitized prostate biopsies. We validated that a diverse set of submitted algorithms reached pathologist-level performance on independent cross-continental cohorts, fully blinded to the algorithm developers. On United States and European external validation sets, the algorithms achieved agreements of 0.862 (quadratically weighted κ, 95% confidence interval (CI), 0.840-0.884) and 0.868 (95% CI, 0.835-0.900) with expert uropathologists. Successful generalization across different patient populations, laboratories and reference standards, achieved by a variety of algorithmic approaches, warrants evaluating AI-based Gleason grading in prospective clinical trials.

Transgenic methods to manipulate CD4 T lymphocytes in vivo via forced expression of TCR transgenes and targeted "knockout" of individual genes by Cre-lox technology are fundamental to modern immunology. However, efforts to scale up functional analysis by modifying expression of larger numbers of genes in T cells ex vivo have proven surprisingly difficult. Early RNA interference experiments achieved successful small RNA transfection by using very high concentrations of short-interfering RNA (siRNA) [1], but primary T cells are generally resistant to standard electroporation, cationic liposome-, and calcium phosphate-mediated transfection methods. Moreover, although viral vectors can successfully introduce DNA fragments of varying length, expression of these constructs in primary T cells is low efficiency and the subcloning process laborious. In this context, the relatively recent discovery of dozens of highly expressed microRNAs (miRNAs) in the immune system provides both an opportunity and a new challenge [2, 3]. How can we query the miRNAome of a cell to assign particular roles to individual miRNAs? Here, we describe an optimized technique for efficient and reproducible transfection of primary mouse CD4 T cells in vitro with synthetic miRNA mimics.

MicroRNAs (miRNAs) are important regulators of cell fate decisions in immune responses. They act by coordinate repression of multiple target genes, a property that we exploited to uncover regulatory networks that govern T helper-2 (Th2) cells. A functional screen of individual miRNAs in primary T cells uncovered multiple miRNAs that inhibited Th2 cell differentiation. Among these were miR-24 and miR-27, miRNAs coexpressed from two genomic clusters, which each functioned independently to limit interleukin-4 (IL-4) production. Mice lacking both clusters in T cells displayed increased Th2 cell responses and tissue pathology in a mouse model of asthma. Gene expression and pathway analyses placed miR-27 upstream of genes known to regulate Th2 cells. They also identified targets not previously associated with Th2 cell biology which regulated IL-4 production in unbiased functional testing. Thus, elucidating the biological function and target repertoire of miR-24 and miR-27 reveals regulators of Th2 cell biology.

Providing prognostic information at the time of cancer diagnosis has important implications for treatment and monitoring. Although cancer staging, histopathological assessment, molecular features, and clinical variables can provide useful prognostic insights, improving risk stratification remains an active research area. We developed a deep learning system (DLS) to predict disease specific survival across 10 cancer types from The Cancer Genome Atlas (TCGA). We used a weakly-supervised approach without pixel-level annotations, and tested three different survival loss functions. The DLS was developed using 9,086 slides from 3,664 cases and evaluated using 3,009 slides from 1,216 cases. In multivariable Cox regression analysis of the combined cohort including all 10 cancers, the DLS was significantly associated with disease specific survival (hazard ratio of 1.58, 95% CI 1.28–1.70, p<0.0001) after adjusting for cancer type, stage, age, and sex. In a per-cancer adjusted subanalysis, the DLS remained a significant predictor of survival in 5 of 10 cancer types. Compared to a baseline model including stage, age, and sex, the c-index of the model demonstrated an absolute 3.7% improvement (95% CI 1.0–6.5) in the combined cohort. Additionally, our models stratified patients within individual cancer stages, particularly stage II (p = 0.025) and stage III (p<0.001). By developing and evaluating prognostic models across multiple cancer types, this work represents one of the most comprehensive studies exploring the direct prediction of clinical outcomes using deep learning and histopathology images. Our analysis demonstrates the potential for this approach to provide significant prognostic information in multiple cancer types, and even within specific pathologic stages. However, given the relatively small number of cases and observed clinical events for a deep learning task of this type, we observed wide confidence intervals for model performance, thus highlighting that future work will benefit from larger datasets assembled for the purposes for survival modeling. [ABSTRACT FROM AUTHOR]

The binding of the RNA polymerase III (pol III) transcription factor TFIIIC to the box A intragenic promoter element of tRNA genes specifies the placement of TFIIIB on upstream-lying DNA. In turn, TFIIIB recruits pol III to the promoter and specifies transcription initiating 17–19 base pairs upstream of box A. The resolution of the pol III transcription apparatus into recombinant TFIIIB, highly purified TFIIIC, and pol III is accompanied by a loss of precision in specifying where transcription initiation occurs due to heterogeneous placement of TFIIIB. In this paper we show that Nhp6a, an abundant high mobility group B (HMGB) family, non-sequence-specific DNA-binding protein in Saccharomyces cerevisiae restores transcriptional initiation fidelity to this highly purified in vitro system. Restoration of initiation fidelity requires the presence of Nhp6a prior to TFIIIB-DNA complex formation. Chemical nuclease footprinting of TFIIIC- and TFIIIB-TFIIIC-DNA complexes reveals that Nhp6a markedly alters the TFIIIC footprint over box A and reduces the size of the TFIIIB footprint on upstream DNA sequence. Analyses of unprocessed tRNAs from yeast lacking Nhp6a and its closely related paralogue Nhp6b demonstrate that Nhp6 is required for transcriptional initiation fidelity of some but not all tRNA genes, in vivo. [ABSTRACT FROM AUTHOR]

Background: Breast cancer management depends on biomarkers including estrogen receptor, progesterone receptor, and human epidermal growth factor receptor 2 (ER/PR/HER2). Though existing scoring systems are widely used and well-validated, they can involve costly preparation and variable interpretation. Additionally, discordances between histology and expected biomarker findings can prompt repeat testing to address biological, interpretative, or technical reasons for unexpected results., Methods: We developed three independent deep learning systems (DLS) to directly predict ER/PR/HER2 status for both focal tissue regions (patches) and slides using hematoxylin-and-eosin-stained (H&E) images as input. Models were trained and evaluated using pathologist annotated slides from three data sources. Areas under the receiver operator characteristic curve (AUCs) were calculated for test sets at both a patch-level (>135 million patches, 181 slides) and slide-level ( n  = 3274 slides, 1249 cases, 37 sites). Interpretability analyses were performed using Testing with Concept Activation Vectors (TCAV), saliency analysis, and pathologist review of clustered patches., Results: The patch-level AUCs are 0.939 (95%CI 0.936-0.941), 0.938 (0.936-0.940), and 0.808 (0.802-0.813) for ER/PR/HER2, respectively. At the slide level, AUCs are 0.86 (95%CI 0.84-0.87), 0.75 (0.73-0.77), and 0.60 (0.56-0.64) for ER/PR/HER2, respectively. Interpretability analyses show known biomarker-histomorphology associations including associations of low-grade and lobular histology with ER/PR positivity, and increased inflammatory infiltrates with triple-negative staining., Conclusions: This study presents rapid breast cancer biomarker estimation from routine H&E slides and builds on prior advances by prioritizing interpretability of computationally learned features in the context of existing pathological knowledge., Competing Interests: Competing interestsThis study was funded by Google LLC and Verily Life Sciences. P.G., R.J., H.W., F.T., M.M., G.S.C., L.H.P., Y.L., C.H.M., D.F.S., and P.-H.C.C. are employees of Google LLC and own Alphabet stock. I.F.-A. and T.B. are consultants of Google LLC. M.T., D.J.D., E.A.R., P.R., N.O., J.H.W., and C.R. declare no competing interests., (© The Author(s) 2021.)

BackgroundDeep learning has the potential to augment the use of chest radiography in clinical radiology, but challenges include poor generalizability, spectrum bias, and difficulty comparing across studies.PurposeTo develop and evaluate deep learning models for chest radiograph interpretation by using radiologist-adjudicated reference standards.Materials and MethodsDeep learning models were developed to detect four findings (pneumothorax, opacity, nodule or mass, and fracture) on frontal chest radiographs. This retrospective study used two data sets. Data set 1 (DS1) consisted of 759 611 images from a multicity hospital network and ChestX-ray14 is a publicly available data set with 112 120 images. Natural language processing and expert review of a subset of images provided labels for 657 954 training images. Test sets consisted of 1818 and 1962 images from DS1 and ChestX-ray14, respectively. Reference standards were defined by radiologist-adjudicated image review. Performance was evaluated by area under the receiver operating characteristic curve analysis, sensitivity, specificity, and positive predictive value. Four radiologists reviewed test set images for performance comparison. Inverse probability weighting was applied to DS1 to account for positive radiograph enrichment and estimate population-level performance.ResultsIn DS1, population-adjusted areas under the receiver operating characteristic curve for pneumothorax, nodule or mass, airspace opacity, and fracture were, respectively, 0.95 (95% confidence interval [CI]: 0.91, 0.99), 0.72 (95% CI: 0.66, 0.77), 0.91 (95% CI: 0.88, 0.93), and 0.86 (95% CI: 0.79, 0.92). With ChestX-ray14, areas under the receiver operating characteristic curve were 0.94 (95% CI: 0.93, 0.96), 0.91 (95% CI: 0.89, 0.93), 0.94 (95% CI: 0.93, 0.95), and 0.81 (95% CI: 0.75, 0.86), respectively.ConclusionExpert-level models for detecting clinically relevant chest radiograph findings were developed for this study by using adjudicated reference standards and with population-level performance estimation. Radiologist-adjudicated labels for 2412 ChestX-ray14 validation set images and 1962 test set images are provided.© RSNA, 2019 Online supplemental material is available for this article. See also the editorial by Chang in this issue.

The past two decades have seen unprecedented advances in the field of oncogenomics. The ongoing characterization of neoplastic tissues through genomic techniques has transformed many aspects of cancer research, diagnosis, and treatment. However, identifying sequence variants with biological and clinical significance is a challenging endeavor. In order to accomplish this task, variants must be annotated and interpreted using various online resources. Data on protein structure, functional prediction, variant frequency in relevant populations, and multiple other factors have been compiled in useful databases for this purpose. Thus, understanding the available online resources for the annotation and interpretation of sequence variants is critical to aid molecular pathologists and researchers working in this space.

Transgenic methods to manipulate CD4 T lymphocytes in vivo via forced expression of TCR transgenes and targeted "knockout" of individual genes by Cre-lox technology are fundamental to modern immunology. However, efforts to scale up functional analysis by modifying expression of larger numbers of genes in T cells ex vivo have proven surprisingly difficult. Early RNA interference experiments achieved successful small RNA transfection by using very high concentrations of short-interfering RNA (siRNA) [1], but primary T cells are generally resistant to standard electroporation, cationic liposome-, and calcium phosphate-mediated transfection methods. Moreover, although viral vectors can successfully introduce DNA fragments of varying length, expression of these constructs in primary T cells is low efficiency and the subcloning process laborious. In this context, the relatively recent discovery of dozens of highly expressed microRNAs (miRNAs) in the immune system provides both an opportunity and a new challenge [2, 3]. How can we query the miRNAome of a cell to assign particular roles to individual miRNAs? Here, we describe an optimized technique for efficient and reproducible transfection of primary mouse CD4 T cells in vitro with synthetic miRNA mimics.

MicroRNAs (miRNAs) are small noncoding RNAs that have recently emerged as critical regulators of gene expression within the immune system. In this study, we used mice with conditional deletion of Dicer and DiGeorge syndrome critical region 8 (Dgcr8) to dissect the roles of miRNAs in NK cell activation, survival, and function during viral infection. We developed a system for deletion of either Dicer or Dgcr8 in peripheral NK cells via drug-induced Cre activity. We found that Dicer- and Dgcr8-deficient NK cells were significantly impaired in survival and turnover, and had impaired function of the ITAM-containing activating NK cell receptors. We further demonstrated that both Dicer- and Dgcr8-dependent pathways were indispensable for the expansion of Ly49H(+) NK cells during mouse cytomegalovirus infection. Our data indicate similar phenotypes for Dicer- and Dgcr8-deficient NK cells, which strongly suggest that these processes are regulated by miRNAs. Thus, our findings indicate a critical role for miRNAs in controlling NK cell homeostasis and effector function, with implications for miRNAs regulating diverse aspects of NK cell biology.