Search

Abstract Liquid biopsies have the potential to revolutionize cancer care through non-invasive early detection of tumors. Developing a robust liquid biopsy test requires collecting high-dimensional data from a large number of blood samples across heterogeneous groups of patients. We propose that the generative capability of variational auto-encoders enables learning a robust and generalizable signature of blood-based biomarkers. In this study, we analyze orphan non-coding RNAs (oncRNAs) from serum samples of 1050 individuals diagnosed with non-small cell lung cancer (NSCLC) at various stages, as well as sex-, age-, and BMI-matched controls. We demonstrate that our multi-task generative AI model, Orion, surpasses commonly used methods in both overall performance and generalizability to held-out datasets. Orion achieves an overall sensitivity of 94% (95% CI: 87%–98%) at 87% (95% CI: 81%–93%) specificity for cancer detection across all stages, outperforming the sensitivity of other methods on held-out validation datasets by more than ~ 30%.

Abstract Despite the importance of informed consent in healthcare, the readability and specificity of consent forms often impede patients’ comprehension. This study investigates the use of GPT-4 to simplify surgical consent forms and introduces an AI-human expert collaborative approach to validate content appropriateness. Consent forms from multiple institutions were assessed for readability and simplified using GPT-4, with pre- and post-simplification readability metrics compared using nonparametric tests. Independent reviews by medical authors and a malpractice defense attorney were conducted. Finally, GPT-4’s potential for generating de novo procedure-specific consent forms was assessed, with forms evaluated using a validated 8-item rubric and expert subspecialty surgeon review. Analysis of 15 academic medical centers’ consent forms revealed significant reductions in average reading time, word rarity, and passive sentence frequency (all P

Background: Patients are increasingly using artificial intelligence (AI) chatbots to seek answers to medical queries. Methods: Ten frequently asked questions in anaesthesia were posed to three AI chatbots: ChatGPT4 (OpenAI), Bard (Google), and Bing Chat (Microsoft). Each chatbot's answers were evaluated in a randomised, blinded order by five residency programme directors from 15 medical institutions in the USA. Three medical content quality categories (accuracy, comprehensiveness, safety) and three communication quality categories (understandability, empathy/respect, and ethics) were scored between 1 and 5 (1 representing worst, 5 representing best). Results: ChatGPT4 and Bard outperformed Bing Chat (median [inter-quartile range] scores: 4 [3–4], 4 [3–4], and 3 [2–4], respectively; P

Abstract Transesophageal echocardiography (TEE) imaging is a vital tool used in the evaluation of complex cardiac pathology and the management of cardiac surgery patients. A key limitation to the application of deep learning strategies to intraoperative and intraprocedural TEE data is the complexity and unstructured nature of these images. In the present study, we developed a deep learning-based, multi-category TEE view classification model that can be used to add structure to intraoperative and intraprocedural TEE imaging data. More specifically, we trained a convolutional neural network (CNN) to predict standardized TEE views using labeled intraoperative and intraprocedural TEE videos from Cedars-Sinai Medical Center (CSMC). We externally validated our model on intraoperative TEE videos from Stanford University Medical Center (SUMC). Accuracy of our model was high across all labeled views. The highest performance was achieved for the Trans-Gastric Left Ventricular Short Axis View (area under the receiver operating curve [AUC] = 0.971 at CSMC, 0.957 at SUMC), the Mid-Esophageal Long Axis View (AUC = 0.954 at CSMC, 0.905 at SUMC), the Mid-Esophageal Aortic Valve Short Axis View (AUC = 0.946 at CSMC, 0.898 at SUMC), and the Mid-Esophageal 4-Chamber View (AUC = 0.939 at CSMC, 0.902 at SUMC). Ultimately, we demonstrate that our deep learning model can accurately classify standardized TEE views, which will facilitate further downstream deep learning analyses for intraoperative and intraprocedural TEE imaging.

Abstract Deep neural networks (DNNs) extract thousands to millions of task-specific features during model training for inference and decision-making. While visualizing these features is critical for comprehending the learning process and improving the performance of the DNNs, existing visualization techniques work only for classification tasks. For regressions, the feature points lie on a high dimensional continuum having an inherently complex shape, making a meaningful visualization of the features intractable. Given that the majority of deep learning applications are regression-oriented, developing a conceptual framework and computational method to reliably visualize the regression features is of great significance. Here, we introduce a manifold discovery and analysis (MDA) method for DNN feature visualization, which involves learning the manifold topology associated with the output and target labels of a DNN. MDA leverages the acquired topological information to preserve the local geometry of the feature space manifold and provides insightful visualizations of the DNN features, highlighting the appropriateness, generalizability, and adversarial robustness of a DNN. The performance and advantages of the MDA approach compared to the existing methods are demonstrated in different deep learning applications.

Abstract Risk prediction models using genetic data have seen increasing traction in genomics. However, most of the polygenic risk models were developed using data from participants with similar (mostly European) ancestry. This can lead to biases in the risk predictors resulting in poor generalization when applied to minority populations and admixed individuals such as African Americans. To address this issue, largely due to the prediction models being biased by the underlying population structure, we propose a deep-learning framework that leverages data from diverse population and disentangles ancestry from the phenotype-relevant information in its representation. The ancestry disentangled representation can be used to build risk predictors that perform better across minority populations. We applied the proposed method to the analysis of Alzheimer’s disease genetics. Comparing with standard linear and nonlinear risk prediction methods, the proposed method substantially improves risk prediction in minority populations, including admixed individuals, without needing self-reported ancestry information.

Abstract The electrocardiogram (ECG) is the most frequently performed cardiovascular diagnostic test, but it is unclear how much information resting ECGs contain about long term cardiovascular risk. Here we report that a deep convolutional neural network can accurately predict the long-term risk of cardiovascular mortality and disease based on a resting ECG alone. Using a large dataset of resting 12-lead ECGs collected at Stanford University Medical Center, we developed SEER, the Stanford Estimator of Electrocardiogram Risk. SEER predicts 5-year cardiovascular mortality with an area under the receiver operator characteristic curve (AUC) of 0.83 in a held-out test set at Stanford, and with AUCs of 0.78 and 0.83 respectively when independently evaluated at Cedars-Sinai Medical Center and Columbia University Irving Medical Center. SEER predicts 5-year atherosclerotic disease (ASCVD) with an AUC of 0.67, similar to the Pooled Cohort Equations for ASCVD Risk, while being only modestly correlated. When used in conjunction with the Pooled Cohort Equations, SEER accurately reclassified 16% of patients from low to moderate risk, uncovering a group with an actual average 9.9% 10-year ASCVD risk who would not have otherwise been indicated for statin therapy. SEER can also predict several other cardiovascular conditions such as heart failure and atrial fibrillation. Using only lead I of the ECG it predicts 5-year cardiovascular mortality with an AUC of 0.80. SEER, used alongside the Pooled Cohort Equations and other risk tools, can substantially improve cardiovascular risk stratification and aid in medical decision making.

Background: Segmenting computed tomography (CT) is crucial in various clinical applications, such as tailoring personalized cardiac ablation for managing cardiac arrhythmias. Automating segmentation through machine learning (ML) is hindered by the necessity for large, labeled training data, which can be challenging to obtain. This article proposes a novel approach for automated, robust labeling using domain knowledge to achieve high-performance segmentation by ML from a small training set. The approach, the domain knowledge-encoding (DOKEN) algorithm, reduces the reliance on large training datasets by encoding cardiac geometry while automatically labeling the training set. The method was validated in a hold-out dataset of CT results from an atrial fibrillation (AF) ablation study. Methods: The DOKEN algorithm parses left atrial (LA) structures, extracts “anatomical knowledge” by leveraging digital LA models (available publicly), and then applies this knowledge to achieve high ML segmentation performance with a small number of training samples. The DOKEN-labeled training set was used to train a nnU-Net deep neural network (DNN) model for segmenting cardiac CT in N = 20 patients. Subsequently, the method was tested in a hold-out set with N = 100 patients (five times larger than training set) who underwent AF ablation. Results: The DOKEN algorithm integrated with the nn-Unet model achieved high segmentation performance with few training samples, with a training to test ratio of 1:5. The Dice score of the DOKEN-enhanced model was 96.7% (IQR: 95.3% to 97.7%), with a median error in surface distance of boundaries of 1.51 mm (IQR: 0.72 to 3.12) and a mean centroid–boundary distance of 1.16 mm (95% CI: −4.57 to 6.89), similar to expert results (r = 0.99; p < 0.001). In digital hearts, the novel DOKEN approach segmented the LA structures with a mean difference for the centroid–boundary distances of −0.27 mm (95% CI: −3.87 to 3.33; r = 0.99; p < 0.0001). Conclusions: The proposed novel domain knowledge-encoding algorithm was able to perform the segmentation of six substructures of the LA, reducing the need for large training data sets. The combination of domain knowledge encoding and a machine learning approach could reduce the dependence of ML on large training datasets and could potentially be applied to AF ablation procedures and extended in the future to other imaging, 3D printing, and data science applications.

Abstract Images depicting dark skin tones are significantly underrepresented in the educational materials used to teach primary care physicians and dermatologists to recognize skin diseases. This could contribute to disparities in skin disease diagnosis across different racial groups. Previously, domain experts have manually assessed textbooks to estimate the diversity in skin images. Manual assessment does not scale to many educational materials and introduces human errors. To automate this process, we present the Skin Tone Analysis for Representation in EDucational materials (STAR-ED) framework, which assesses skin tone representation in medical education materials using machine learning. Given a document (e.g., a textbook in .pdf), STAR-ED applies content parsing to extract text, images, and table entities in a structured format. Next, it identifies images containing skin, segments the skin-containing portions of those images, and estimates the skin tone using machine learning. STAR-ED was developed using the Fitzpatrick17k dataset. We then externally tested STAR-ED on four commonly used medical textbooks. Results show strong performance in detecting skin images (0.96 ± 0.02 AUROC and 0.90 ± 0.06 F1 score) and classifying skin tones (0.87 ± 0.01 AUROC and 0.91 ± 0.00 F1 score). STAR-ED quantifies the imbalanced representation of skin tones in four medical textbooks: brown and black skin tones (Fitzpatrick V-VI) images constitute only 10.5% of all skin images. We envision this technology as a tool for medical educators, publishers, and practitioners to assess skin tone diversity in their educational materials.

In recent years, there has been a dramatic rise in interest in retrosynthesis prediction with artificial intelligence (AI) techniques. Unlike conventional retrosynthesis prediction performed by chemists and by rule-based expert systems, AI-driven retrosynthesis prediction automatically learns chemistry knowledge from off-the-shelf experimental datasets to predict reactions and retrosynthesis routes. This provides an opportunity to address many conventional challenges, including heavy reliance on extensive expertise, the sub-optimality of routes, and prohibitive computational cost. This review describes the current landscape of AI-driven retrosynthesis prediction. We first discuss formal definitions of the retrosynthesis problem and review the outstanding research challenges therein. We then review the related AI techniques and recent progress that enable retrosynthesis prediction. Moreover, we propose a novel landscape that provides a comprehensive categorization of different retrosynthesis prediction components and survey how AI reshapes each component. We conclude by discussing promising areas for future research.

Abstract Background Undiagnosed chronic kidney disease (CKD) is a common and usually asymptomatic disorder that causes a high burden of morbidity and early mortality worldwide. We developed a deep learning model for CKD screening from routinely acquired ECGs. Methods We collected data from a primary cohort with 111,370 patients which had 247,655 ECGs between 2005 and 2019. Using this data, we developed, trained, validated, and tested a deep learning model to predict whether an ECG was taken within one year of the patient receiving a CKD diagnosis. The model was additionally validated using an external cohort from another healthcare system which had 312,145 patients with 896,620 ECGs between 2005 and 2018. Results Using 12-lead ECG waveforms, our deep learning algorithm achieves discrimination for CKD of any stage with an AUC of 0.767 (95% CI 0.760–0.773) in a held-out test set and an AUC of 0.709 (0.708–0.710) in the external cohort. Our 12-lead ECG-based model performance is consistent across the severity of CKD, with an AUC of 0.753 (0.735–0.770) for mild CKD, AUC of 0.759 (0.750–0.767) for moderate-severe CKD, and an AUC of 0.783 (0.773–0.793) for ESRD. In patients under 60 years old, our model achieves high performance in detecting any stage CKD with both 12-lead (AUC 0.843 [0.836–0.852]) and 1-lead ECG waveform (0.824 [0.815–0.832]). Conclusions Our deep learning algorithm is able to detect CKD using ECG waveforms, with stronger performance in younger patients and more severe CKD stages. This ECG algorithm has the potential to augment screening for CKD.

BackgroundSegmentation of computed tomography (CT) is important for many clinical procedures including personalized cardiac ablation for the management of cardiac arrhythmias. While segmentation can be automated by machine learning (ML), it is limited by the need for large, labeled training data that may be difficult to obtain. We set out to combine ML of cardiac CT with domain knowledge, which reduces the need for large training datasets by encoding cardiac geometry, which we then tested in independent datasets and in a prospective study of atrial fibrillation (AF) ablation.MethodsWe mathematically represented atrial anatomy with simple geometric shapes and derived a model to parse cardiac structures in a small set of N = 6 digital hearts. The model, termed “virtual dissection,” was used to train ML to segment cardiac CT in N = 20 patients, then tested in independent datasets and in a prospective study.ResultsIn independent test cohorts (N = 160) from 2 Institutions with different CT scanners, atrial structures were accurately segmented with Dice scores of 96.7% in internal (IQR: 95.3%–97.7%) and 93.5% in external (IQR: 91.9%–94.7%) test data, with good agreement with experts (r = 0.99; p

Dimension reduction is an indispensable part of modern data science, and many algorithms have been developed. Here, the authors develop a theoretically justified, simple to use and reliable spectral method to assess and combine multiple dimension reduction visualizations of a given dataset from diverse algorithms.

Abstract Spontaneous tumors in canines share significant genetic and histological similarities with human tumors, positioning them as valuable models to guide drug development. However, current translational studies have limited real world evidence as cancer outcomes are dispersed across veterinary clinics and genomic tests are rarely performed on dogs. In this study, we aim to expand the value of canine models by systematically characterizing genetic mutations in tumors and their response to targeted treatments. In total, we collect and analyze survival outcomes for 2119 tumor-bearing dogs and the prognostic effect of genomic alterations in a subset of 1108 dogs. Our analysis identifies prognostic concordance between canines and humans in several key oncogenes, including TP53 and PIK3CA. We also find that several targeted treatments designed for humans are associated with a positive prognosis when used to treat canine tumors with specific genomic alterations, underscoring the value of canine models in advancing drug discovery for personalized oncology.

Abstract Prostate cancer is the most frequent cancer in men and a leading cause of cancer death. Determining a patient’s optimal therapy is a challenge, where oncologists must select a therapy with the highest likelihood of success and the lowest likelihood of toxicity. International standards for prognostication rely on non-specific and semi-quantitative tools, commonly leading to over- and under-treatment. Tissue-based molecular biomarkers have attempted to address this, but most have limited validation in prospective randomized trials and expensive processing costs, posing substantial barriers to widespread adoption. There remains a significant need for accurate and scalable tools to support therapy personalization. Here we demonstrate prostate cancer therapy personalization by predicting long-term, clinically relevant outcomes using a multimodal deep learning architecture and train models using clinical data and digital histopathology from prostate biopsies. We train and validate models using five phase III randomized trials conducted across hundreds of clinical centers. Histopathological data was available for 5654 of 7764 randomized patients (71%) with a median follow-up of 11.4 years. Compared to the most common risk-stratification tool—risk groups developed by the National Cancer Center Network (NCCN)—our models have superior discriminatory performance across all endpoints, ranging from 9.2% to 14.6% relative improvement in a held-out validation set. This artificial intelligence-based tool improves prognostication over standard tools and allows oncologists to computationally predict the likeliest outcomes of specific patients to determine optimal treatment. Outfitted with digital scanners and internet access, any clinic could offer such capabilities, enabling global access to therapy personalization.

Abstract Background Transparent and accessible reporting of COVID-19 data is critical for public health efforts. Each Indian state has its own mechanism for reporting COVID-19 data, and the quality of their reporting has not been systematically evaluated. We present a comprehensive assessment of the quality of COVID-19 data reporting done by the Indian state governments between 19 May and 1 June, 2020. Methods We designed a semi-quantitative framework with 45 indicators to assess the quality of COVID-19 data reporting. The framework captures four key aspects of public health data reporting – availability, accessibility, granularity, and privacy. We used this framework to calculate a COVID-19 Data Reporting Score (CDRS, ranging from 0–1) for each state. Results Our results indicate a large disparity in the quality of COVID-19 data reporting across India. CDRS varies from 0.61 (good) in Karnataka to 0.0 (poor) in Bihar and Uttar Pradesh, with a median value of 0.26. Ten states do not report data stratified by age, gender, comorbidities or districts. Only ten states provide trend graphics for COVID-19 data. In addition, we identify that Punjab and Chandigarh compromised the privacy of individuals under quarantine by publicly releasing their personally identifiable information. The CDRS is positively associated with the state’s sustainable development index for good health and well-being (Pearson correlation: r=0.630,p=0.0003). Conclusions Our assessment informs the public health efforts in India and serves as a guideline for pandemic data reporting. The disparity in CDRS highlights three important findings at the national, state, and individual level. At the national level, it shows the lack of a unified framework for reporting COVID-19 data in India, and highlights the need for a central agency to monitor or audit the quality of data reporting done by the states. Without a unified framework, it is difficult to aggregate the data from different states, gain insights, and coordinate an effective nationwide response to the pandemic. Moreover, it reflects the inadequacy in coordination or sharing of resources among the states. The disparate reporting score also reflects inequality in individual access to public health information and privacy protection based on the state of residence.

Abstract The reliability of machine learning models can be compromised when trained on low quality data. Many large-scale medical imaging datasets contain low quality labels extracted from sources such as medical reports. Moreover, images within a dataset may have heterogeneous quality due to artifacts and biases arising from equipment or measurement errors. Therefore, algorithms that can automatically identify low quality data are highly desired. In this study, we used data Shapley, a data valuation metric, to quantify the value of training data to the performance of a pneumonia detection algorithm in a large chest X-ray dataset. We characterized the effectiveness of data Shapley in identifying low quality versus valuable data for pneumonia detection. We found that removing training data with high Shapley values decreased the pneumonia detection performance, whereas removing data with low Shapley values improved the model performance. Furthermore, there were more mislabeled examples in low Shapley value data and more true pneumonia cases in high Shapley value data. Our results suggest that low Shapley value indicates mislabeled or poor quality images, whereas high Shapley value indicates data that are valuable for pneumonia detection. Our method can serve as a framework for using data Shapley to denoise large-scale medical imaging datasets.

Abstract Genome-wide pooled CRISPR-Cas-mediated knockout, activation, and repression screens are powerful tools for functional genomic investigations. Despite their increasing importance, there is currently little guidance on how to design and analyze CRISPR-pooled screens. Here, we provide a review of the commonly used algorithms in the computational analysis of pooled CRISPR screens. We develop a comprehensive simulation framework to benchmark and compare the performance of these algorithms using both synthetic and real datasets. Our findings inform parameter choices of CRISPR screens and provide guidance to researchers on the design and analysis of pooled CRISPR screens.

India is among the top three countries in the world both in COVID-19 case and death counts. With the pandemic far from over, timely, transparent, and accessible reporting of COVID-19 data continues to be critical for India's pandemic efforts. We systematically analyze the quality of reporting of COVID-19 data in over one hundred government platforms (web and mobile) from India. Our analyses reveal a lack of granular data in the reporting of COVID-19 surveillance, vaccination, and vacant bed availability. As of 5 June 2021, age and gender distribution are available for less than 22% of cases and deaths, and comorbidity distribution is available for less than 30% of deaths. Amid rising concerns of undercounting cases and deaths in India, our results highlight a patchy reporting of granular data even among the reported cases and deaths. Furthermore, total vaccination stratified by healthcare workers, frontline workers, and age brackets is reported by only 14 out of India's 36 subnationals (states and union territories). There is no reporting of adverse events following immunization by vaccine and event type. By showing what, where, and how much data is missing, we highlight the need for a more responsible and transparent reporting of granular COVID-19 data in India.

Background In recent years, interest has grown in whether and to what extent demographic diversity sparks discovery and innovation in research. At the same time, topic modeling has been employed to discover differences in what women and men write about. This study engages these two strands of scholarship to explore associations between changing researcher demographics and research questions asked in the discipline of history. Specifically, we analyze developments in history as women entered the field. Methods We focus on author gender in diachronic analysis of history dissertations from 1980 (when online data is first available) to 2015 and a select set of general history journals from 1950 to 2015. We use correlated topic modeling and network visualizations to map developments in research agendas over time and to examine how women and men have contributed to these developments. Results Our summary snapshot of aggregate interests of women and men for the period 1950 to 2015 identifies new topics associated with women authors: gender and women’s history, body history, family and households, consumption and consumerism, and sexuality. Diachronic analysis demonstrates that while women pioneered topics such as gender and women’s history or the history of sexuality, these topics broaden over time to become methodological frameworks that historians widely embraced and that changed in interesting ways as men engaged with them. Our analysis of history dissertations surface correlations between advisor/advisee gender pairings and choice of dissertation topic. Conclusions Overall, this quantitative longitudinal study suggests that the growth in women historians has coincided with the broadening of research agendas and an increased sensitivity to new topics and methodologies in the field.

Background: Despite a 2010 Kenyan constitutional amendment limiting members of elected public bodies to < two-thirds of the same gender, only 22 percent of the 12th Parliament members inaugurated in 2017 were women. Investigating gender bias in the media is a useful tool for understanding socio-cultural barriers to implementing legislation for gender equality. Natural language processing (NLP) methods, such as word embedding and sentiment analysis, can efficiently quantify media biases at a scope previously unavailable in the social sciences.Methods: We trained GloVe and word2vec word embeddings on text from 1998 to 2019 from Kenya’s Daily Nation newspaper. We measured gender bias in these embeddings and used sentiment analysis to predict quantitative sentiment scores for sentences surrounding female leader names compared to male leader names.Results: Bias in leadership words for men and women measured from Daily Nation word embeddings corresponded to temporal trends in men and women’s participation in political leadership (i.e., parliamentary seats) using GloVe (correlation 0.8936, p = 0.0067, r2 = 0.799) and word2vec (correlation 0.844, p = 0.0169, r2 = 0.712) algorithms. Women continue to be associated with domestic terms while men continue to be associated with influence terms, for both regular gender words and female and male political leaders’ names. Male words (e.g., he, him, man) were mentioned 1.84 million more times than female words from 1998 to 2019. Sentiment analysis showed an increase in relative negative sentiment associated with female leaders (p = 0.0152) and an increase in positive sentiment associated with male leaders over time (p = 0.0216).Conclusion: Natural language processing is a powerful method for gaining insights into and quantifying trends in gender biases and sentiment in news media. We found evidence of improvement in gender equality but also a backlash from increased female representation in high-level governmental leadership.

Side information in addition to the p-values is often available in modern applications of multiple hypothesis testing. Here, the authors develop AdaFDR, a new statistical method for multiple hypothesis testing that adaptively learns the decision threshold and amplifies the discovery power.

Artificial Intelligence (AI) can potentially impact many aspects of human health, from basic research discovery to individual health assessment. It is critical that these advances in technology broadly benefit diverse populations from around the world. This can be challenging because AI algorithms are often developed on non-representative samples and evaluated based on narrow metrics. Here we outline key challenges to biomedical AI in outcome design, data collection and technology evaluation, and use examples from precision health to illustrate how bias and health disparity may arise in each stage. We then suggest both short term approaches—more diverse data collection and AI monitoring—and longer term structural changes in funding, publications, and education to address these challenges.

Aging is associated with complex molecular and cellular processes that are poorly understood. Here we leveraged the Tabula Muris Senis single-cell RNA-seq data set to systematically characterize gene expression changes during aging across diverse cell types in the mouse. We identified aging-dependent genes in 76 tissue-cell types from 23 tissues and characterized both shared and tissue-cell-specific aging behaviors. We found that the aging-related genes shared by multiple tissue-cell types also change their expression congruently in the same direction during aging in most tissue-cell types, suggesting a coordinated global aging behavior at the organismal level. Scoring cells based on these shared aging genes allowed us to contrast the aging status of different tissues and cell types from a transcriptomic perspective. In addition, we identified genes that exhibit age-related expression changes specific to each functional category of tissue-cell types. Altogether, our analyses provide one of the most comprehensive and systematic characterizations of the molecular signatures of aging across diverse tissue-cell types in a mammalian system.

Abstract Introduction Positron emission tomography (PET) imaging targeting neurofibrillary tau tangles is increasingly used in the study of Alzheimer's disease (AD), but its utility may be limited by conventional quantitative or qualitative evaluation techniques in earlier disease states. Convolutional neural networks (CNNs) are effective in learning spatial patterns for image classification. Methods 18F‐MK6240 (n = 320) and AV‐1451 (n = 446) PET images were pooled from multiple studies. We performed iterations with differing permutations of radioligands, heuristics, and architectures. Performance was compared to a standard region of interest (ROI)‐based approach on prediction of memory impairment. We visualized attention of the network to illustrate decision making. Results Overall, models had high accuracy (> 80%) with good average sensitivity and specificity (75% and 82%, respectively), and had comparable or higher accuracy to the ROI standard. Visualizations of model attention highlight known characteristics of tau radioligand binding. Discussion CNNs could improve tau PET's role in early disease and extend the utility of tau PET across generations of radioligands.

Dimensionality reduction and visualization methods lack a principled way of comparing multiple datasets. Here, Abid et al. introduce contrastive PCA, which identifies low-dimensional structures enriched in one dataset compared to another and enables visualization of dataset-specific patterns.

Accurate estimations of the frequency distribution of rare variants are needed to quantify the discovery power and guide large-scale human sequencing projects. This study describes an algorithm called UnseenEst to estimate the distribution of genetic variations using tens of thousands of exomes.

Malaria parasites (Plasmodium spp.) and related apicomplexan pathogens contain a nonphotosynthetic plastid called the apicoplast. Derived from an unusual secondary eukaryote-eukaryote endosymbiosis, the apicoplast is a fascinating organelle whose function and biogenesis rely on a complex amalgamation of bacterial and algal pathways. Because these pathways are distinct from the human host, the apicoplast is an excellent source of novel antimalarial targets. Despite its biomedical importance and evolutionary significance, the absence of a reliable apicoplast proteome has limited most studies to the handful of pathways identified by homology to bacteria or primary chloroplasts, precluding our ability to study the most novel apicoplast pathways. Here, we combine proximity biotinylation-based proteomics (BioID) and a new machine learning algorithm to generate a high-confidence apicoplast proteome consisting of 346 proteins. Critically, the high accuracy of this proteome significantly outperforms previous prediction-based methods and extends beyond other BioID studies of unique parasite compartments. Half of identified proteins have unknown function, and 77% are predicted to be important for normal blood-stage growth. We validate the apicoplast localization of a subset of novel proteins and show that an ATP-binding cassette protein ABCF1 is essential for blood-stage survival and plays a previously unknown role in apicoplast biogenesis. These findings indicate critical organellar functions for newly discovered apicoplast proteins. The apicoplast proteome will be an important resource for elucidating unique pathways derived from secondary endosymbiosis and prioritizing antimalarial drug targets.