In coming to understand the world-in learning concepts, acquiring language, and grasping causal relations-our minds make inferences that appear to go far beyond the data available. How do we do it? This review describes recent approaches to reverse-engineering human learning and cognitive development and, in parallel, engineering more humanlike machine learning systems. Computational models that perform probabilistic inference over hierarchies of flexibly structured representations can address some of the deepest questions about the nature and origins of human thought: How does abstract knowledge guide learning and reasoning from sparse data? What forms does our knowledge take, across different domains and tasks? And how is that abstract knowledge itself acquired? <<<

V Gopalakrishnan, C N Spencer, L Nezi, A Reuben, M C Andrews, T V Karpinets, P A Prieto, D Vicente, K Hoffman, S C Wei, A P Cogdill, L Zhao, C W Hudgens, D S Hutchinson, T Manzo, M Petaccia de Macedo, T Cotechini, T Kumar, W S Chen, S M Reddy, R Szczepaniak Sloane, J Galloway-Pena, H Jiang, P L Chen, E J Shpall, K Rezvani, A M Alousi, R F Chemaly, S Shelburne, L M Vence, P C Okhuysen, V B Jensen, A G Swennes, F McAllister, E Marcelo Riquelme Sanchez, Y Zhang, E Le Chatelier, L Zitvogel, N Pons, J L Austin-Breneman, L E Haydu, E M Burton, J M Gardner, E Sirmans, J Hu, A J Lazar, T Tsujikawa, A Diab, H Tawbi, I C Glitza, W J Hwu, S P Patel, S E Woodman, R N Amaria, M A Davies, J E Gershenwald, P Hwu, J E Lee, J Zhang, L M Coussens, Z A Cooper, P A Futreal, C R Daniel, N J Ajami, J F Petrosino, M T Tetzlaff, P Sharma, J P Allison, R R Jenq, J A Wargo <<<

Preclinical mouse models suggest that the gut microbiome modulates tumor response to checkpoint blockade immunotherapy; however, this has not been well-characterized in human cancer patients. Here we examined the oral and gut microbiome of melanoma patients undergoing anti-programmed cell death 1 protein (PD-1) immunotherapy ( = 112). Significant differences were observed in the diversity and composition of the patient gut microbiome of responders versus nonresponders. Analysis of patient fecal microbiome samples ( = 43, 30 responders, 13 nonresponders) showed significantly higher alpha diversity ( < 0.01) and relative abundance of bacteria of the Ruminococcaceae family ( < 0.01) in responding patients. Metagenomic studies revealed functional differences in gut bacteria in responders, including enrichment of anabolic pathways. Immune profiling suggested enhanced systemic and antitumor immunity in responding patients with a favorable gut microbiome as well as in germ-free mice receiving fecal transplants from responding patients. Together, these data have important implications for the treatment of melanoma patients with immune checkpoint inhibitors. <<<

Junyue Cao, Malte Spielmann, Xiaojie Qiu, Xingfan Huang, Daniel M Ibrahim, Andrew J Hill, Fan Zhang, Stefan Mundlos, Lena Christiansen, Frank J Steemers, Cole Trapnell, Jay Shendure <<<

Mammalian organogenesis is a remarkable process. Within a short timeframe, the cells of the three germ layers transform into an embryo that includes most of the major internal and external organs. Here we investigate the transcriptional dynamics of mouse organogenesis at single-cell resolution. Using single-cell combinatorial indexing, we profiled the transcriptomes of around 2 million cells derived from 61 embryos staged between 9.5 and 13.5 days of gestation, in a single experiment. The resulting 'mouse organogenesis cell atlas' (MOCA) provides a global view of developmental processes during this critical window. We use Monocle 3 to identify hundreds of cell types and 56 trajectories, many of which are detected only because of the depth of cellular coverage, and collectively define thousands of corresponding marker genes. We explore the dynamics of gene expression within cell types and trajectories over time, including focused analyses of the apical ectodermal ridge, limb mesenchyme and skeletal muscle. <<<

Benoît Schmauch, Alberto Romagnoni, Elodie Pronier, Charlie Saillard, Pascale Maillé, Julien Calderaro, Aurélie Kamoun, Meriem Sefta, Sylvain Toldo, Mikhail Zaslavskiy, Thomas Clozel, Matahi Moarii, Pierre Courtiol, Gilles Wainrib <<<

Deep learning methods for digital pathology analysis are an effective way to address multiple clinical questions, from diagnosis to prediction of treatment outcomes. These methods have also been used to predict gene mutations from pathology images, but no comprehensive evaluation of their potential for extracting molecular features from histology slides has yet been performed. We show that HE2RNA, a model based on the integration of multiple data modes, can be trained to systematically predict RNA-Seq profiles from whole-slide images alone, without expert annotation. Through its interpretable design, HE2RNA provides virtual spatialization of gene expression, as validated by CD3- and CD20-staining on an independent dataset. The transcriptomic representation learned by HE2RNA can also be transferred on other datasets, even of small size, to increase prediction performance for specific molecular phenotypes. We illustrate the use of this approach in clinical diagnosis purposes such as the identification of tumors with microsatellite instability. <<<

Uncertainty estimation in medical image registration enables surgeons to evaluate the operative risk based on the trustworthiness of the registered image data thus of paramount importance for practical clinical applications. Despite the recent promising results obtained with deep unsupervised learning-based registration methods, reasoning about uncertainty of unsupervised registration models remains largely unexplored. In this work, we propose a predictive module to learn the registration and uncertainty in correspondence simultaneously. Our framework introduces empirical randomness and registration error based uncertainty prediction. We systematically assess the performances on two MRI datasets with different ensemble paradigms. Experimental results highlight that our proposed framework significantly improves the registration accuracy and uncertainty compared with the baseline. <<<

In this work, we propose a theoretical framework based on maximum profile likelihood for pairwise and groupwise registration. By an asymptotic analysis, we demonstrate that maximum profile likelihood registration minimizes an upper bound on the joint entropy of the distribution that generates the joint image data. Further, we derive the congealing method for groupwise registration by optimizing the profile likelihood in closed form, and using coordinate ascent, or iterative model refinement. We also describe a method for feature based registration in the same framework and demonstrate it on groupwise tractographic registration. In the second part of the article, we propose an approach to deep metric registration that implements maximum likelihood registration using deep discriminative classifiers. We show further that this approach can be used for maximum profile likelihood registration to discharge the need for well-registered training data, using iterative model refinement. We demonstrate that the method succeeds on a challenging registration problem where the standard mutual information approach does not perform well. <<<

PURPOSE: The accuracy of minimally invasive, intracranial neurosurgery can be challenged by deformation of brain tissue - e.g., up to 10 mm due to egress of cerebrospinal fluid during neuroendoscopic approach. We report an unsupervised, deep learning-based registration framework to resolve such deformations between preoperative MR and intraoperative CT with fast runtime for neurosurgical guidance.METHOD: The framework incorporates subnetworks for MR and CT image synthesis with a dual-channel registration subnetwork (with synthesis uncertainty providing spatially varying weights on the dual-channel loss) to estimate a diffeomorphic deformation field from both the MR and CT channels. An end-to-end training is proposed that jointly optimizes both the synthesis and registration subnetworks. The proposed framework was investigated using three datasets: (1) paired MR/CT with simulated deformations; (2) paired MR/CT with real deformations; and (3) a neurosurgery dataset with real deformation. Two state-of-the-art methods (Symmetric Normalization and VoxelMorph) were implemented as a basis of comparison, and variations in the proposed dual-channel network were investigated, including single-channel registration, fusion without uncertainty weighting, and conventional sequential training of the synthesis and registration subnetworks.RESULTS: The proposed method achieved: (1) Dice coefficient = 0.82±0.07 and TRE = 1.2 ± 0.6 mm on paired MR/CT with simulated deformations; (2) Dice coefficient = 0.83 ± 0.07 and TRE = 1.4 ± 0.7 mm on paired MR/CT with real deformations; and (3) Dice = 0.79 ± 0.13 and TRE = 1.6 ± 1.0 mm on the neurosurgery dataset with real deformations. The dual-channel registration with uncertainty weighting demonstrated superior performance (e.g., TRE = 1.2 ± 0.6 mm) compared to single-channel registration (TRE = 1.6 ± 1.0 mm, p < 0.05 for CT channel and TRE = 1.3 ± 0.7 mm for MR channel) and dual-channel registration without uncertainty weighting (TRE = 1.4 ± 0.8 mm, p < 0.05). End-to-end training of the synthesis and registration subnetworks also improved performance compared to the conventional sequential training strategy (TRE = 1.3 ± 0.6 mm). Registration runtime with the proposed network was ∼3 s.CONCLUSION: The deformable registration framework based on dual-channel MR/CT registration with spatially varying weights and end-to-end training achieved geometric accuracy and runtime that was superior to state-of-the-art baseline methods and various ablations of the proposed network. The accuracy and runtime of the method may be compatible with the requirements of high-precision neurosurgery. <<<

This paper addresses the scalability challenge of architecture search by formulating the task in a differentiable manner. Unlike conventional approaches of applying evolution or reinforcement learning over a discrete and non-differentiable search space, our method is based on the continuous relaxation of the architecture representation, allowing efficient search of the architecture using gradient descent. Extensive experiments on CIFAR-10, ImageNet, Penn Treebank and WikiText-2 show that our algorithm excels in discovering high-performance convolutional architectures for image classification and recurrent architectures for language modeling, while being orders of magnitude faster than state-of-the-art non-differentiable techniques. Our implementation has been made publicly available to facilitate further research on efficient architecture search algorithms. <<<

The expanding scale and inherent complexity of biological data have encouraged a growing use of machine learning in biology to build informative and predictive models of the underlying biological processes. All machine learning techniques fit models to data; however, the specific methods are quite varied and can at first glance seem bewildering. In this Review, we aim to provide readers with a gentle introduction to a few key machine learning techniques, including the most recently developed and widely used techniques involving deep neural networks. We describe how different techniques may be suited to specific types of biological data, and also discuss some best practices and points to consider when one is embarking on experiments involving machine learning. Some emerging directions in machine learning methodology are also discussed. <<<

The translational challenges in the field of precision oncology are in part related to the biological complexity and diversity of this disease. Technological advances in genomics have facilitated large sequencing efforts and discoveries that have further supported this notion. In this review, we reflect on the impact of these discoveries on our understanding of several concepts: cancer initiation, cancer prevention, early detection, adjuvant therapy and minimal residual disease monitoring, cancer drug resistance, and cancer evolution in metastasis. We discuss key areas of focus for improving cancer outcomes, from biological insights to clinical application, and suggest where the development of these technologies will lead us. Finally, we discuss practical challenges to the wider adoption of molecular profiling in the clinic and the need for robust translational infrastructure. <<<

Rui Tao, Yanhong Wang, Yun Hu, Yaoge Jiao, Lifang Zhou, Lurong Jiang, Li Li, Xingyu He, Min Li, Yamei Yu, Qiang Chen, Shaohua Yao <<<

Large scale genomic aberrations including duplication, deletion, translocation, and other structural changes are the cause of a subtype of hereditary genetic disorders and contribute to onset or progress of cancer. The current prime editor, PE2, consisting of Cas9-nickase and reverse transcriptase enables efficient editing of genomic deletion and insertion, however, at small scale. Here, we designed a novel prime editor by fusing reverse transcriptase (RT) to nuclease wild-type Cas9 (WT-PE) to edit large genomic fragment. WT-PE system simultaneously introduced a double strand break (DSB) and a single 3' extended flap in the target site. Coupled with paired prime editing guide RNAs (pegRNAs) that have complementary sequences in their 3' terminus while target different genomic regions, WT-PE produced bi-directional prime editing, which enabled efficient and versatile large-scale genome editing, including large fragment deletion up to 16.8 megabase (Mb) pairs and chromosomal translocation. Therefore, our WT-PE system has great potential to model or treat diseases related to large-fragment aberrations. <<<

White matter hyperintensities (WMH) of presumed vascular origin are frequently found in MRIs of healthy older adults. WMH are also associated with aging and cognitive decline. Here, we compared and validated three algorithms for WMH extraction: FreeSurfer (T1w), UBO Detector (T1w + FLAIR), and FSL's Brain Intensity AbNormality Classification Algorithm (BIANCA; T1w + FLAIR) using a longitudinal dataset comprising MRI data of cognitively healthy older adults (baseline N = 231, age range 64-87 years). As reference we manually segmented WMH in T1w, three-dimensional (3D) FLAIR, and two-dimensional (2D) FLAIR images which were used to assess the segmentation accuracy of the different automated algorithms. Further, we assessed the relationships of WMH volumes provided by the algorithms with Fazekas scores and age. FreeSurfer underestimated the WMH volumes and scored worst in Dice Similarity Coefficient (DSC = 0.434) but its WMH volumes strongly correlated with the Fazekas scores (r = 0.73). BIANCA accomplished the highest DSC (0.602) in 3D FLAIR images. However, the relations with the Fazekas scores were only moderate, especially in the 2D FLAIR images (r = 0.41), and many outlier WMH volumes were detected when exploring within-person trajectories (2D FLAIR: ~30%). UBO Detector performed similarly to BIANCA in DSC with both modalities and reached the best DSC in 2D FLAIR (0.531) without requiring a tailored training dataset. In addition, it achieved very high associations with the Fazekas scores (2D FLAIR: r = 0.80). In summary, our results emphasize the importance of carefully contemplating the choice of the WMH segmentation algorithm and MR-modality. <<<

Inferring single-cell compositions and their contributions to global gene expression changes from bulk RNA sequencing (RNA-seq) datasets is a major challenge in oncology. Here we develop Bayesian cell proportion reconstruction inferred using statistical marginalization (BayesPrism), a Bayesian method to predict cellular composition and gene expression in individual cell types from bulk RNA-seq, using patient-derived, scRNA-seq as prior information. We conduct integrative analyses in primary glioblastoma, head and neck squamous cell carcinoma and skin cutaneous melanoma to correlate cell type composition with clinical outcomes across tumor types, and explore spatial heterogeneity in malignant and nonmalignant cell states. We refine current cancer subtypes using gene expression annotation after exclusion of confounding nonmalignant cells. Finally, we identify genes whose expression in malignant cells correlates with macrophage infiltration, T cells, fibroblasts and endothelial cells across multiple tumor types. Our work introduces a new lens to accurately infer cellular composition and expression in large cohorts of bulk RNA-seq data. <<<

A central problem in machine learning involves modeling complex data-sets using highly flexible families of probability distributions in which learning, sampling, inference, and evaluation are still analytically or computationally tractable. Here, we develop an approach that simultaneously achieves both flexibility and tractability. The essential idea, inspired by non-equilibrium statistical physics, is to systematically and slowly destroy structure in a data distribution through an iterative forward diffusion process. We then learn a reverse diffusion process that restores structure in data, yielding a highly flexible and tractable generative model of the data. This approach allows us to rapidly learn, sample from, and evaluate probabilities in deep generative models with thousands of layers or time steps, as well as to compute conditional and posterior probabilities under the learned model. We additionally release an open source reference implementation of the algorithm. <<<

It is shown that a long, near microsecond, atomistic simulation can shed some light upon the dynamical processes occurring in a lipid bilayer. The analysis focuses on reorientational dynamics of the chains and lateral diffusion of lipids. It is shown that the reorientational correlation functions exhibits an algebraic decay (rather than exponential) for several orders of magnitude in time. The calculated nuclear magnetic resonance relaxation rates agree with experiments for carbons at the C7 position while there are some differences for C3. Lateral diffusion can be divided into two stages. In a first stage occurring at short times, t<5 ns, the center of mass of the lipid moves due to conformational changes of the chains while the headgroup position remains relatively fixed. In this stage, the center of mass can move up to approximately 0.8 nm. The fitted short-time diffusion coefficient is D(1)=13 x 10(-7) cm(2) s(-1) On a longer time scale, the diffusion coefficient becomes D(2)=0.79 x 10(-7) cm(2) s(-1). <<<

Neoantigen vaccines aiming to induce tumor-specific T cell responses have achieved promising antitumor effects in early clinical trials. However, the underlying mechanism regarding response or resistance to this treatment is unclear. Here we observe that neoantigen vaccine-generated T cells can synergize with the immune checkpoint blockade for effective tumor control. Specifically, we performed single-cell sequencing on over 100,000 T cells and uncovered that combined therapy induces an antigen-specific CD8 T cell population with active chemokine signaling (Cxcr3/Ccl5), lower co-inhibitory receptor expression (Lag3/Havcr2) and higher cytotoxicity (Fasl/Gzma). Furthermore, generation of neoantigen-specific T cells in the draining lymph node is required for combination treatment. Signature genes of this unique population are associated with T cell clonal frequency and better survival in humans. Our study profiles the dynamics of tumor-infiltrating T cells during neoantigen vaccine and immune checkpoint blockade treatments and high-dimensionally identifies neoantigen-reactive T cell signatures for future development of therapeutic strategies. <<<

Biyuan Luo, Fang Ma, Hao Liu, Jixiong Hu, Le Rao, Chun Liu, Yongfang Jiang, Shuyu Kuangzeng, Xuan Lin, Chenyang Wang, Yiyu Lei, Zhongzhou Si, Guangshun Chen, Ning Zhou, Chengbai Liang, Fangqing Jiang, Fenge Liu, Weidong Dai, Wei Liu, Yawen Gao, Zhihong Li, Xi Li, Guangyu Zhou, Bingsi Li, Zhihong Zhang, Weiqi Nian, Lihua Luo, Xianling Liu <<<

BACKGROUND: Aberrant DNA methylation may offer opportunities in revolutionizing cancer screening and diagnosis. We sought to identify a non-invasive DNA methylation-based screening approach using cell-free DNA (cfDNA) for early detection of hepatocellular carcinoma (HCC).METHODS: Differentially, DNA methylation blocks were determined by comparing methylation profiles of biopsy-proven HCC, liver cirrhosis, and normal tissue samples with high throughput DNA bisulfite sequencing. A multi-layer HCC screening model was subsequently constructed based on tissue-derived differentially methylated blocks (DMBs). This model was tested in a cohort consisting of 120 HCC, 92 liver cirrhotic, and 290 healthy plasma samples including 65 hepatitis B surface antigen-seropositive (HBsAg+) samples, independently validated in a cohort consisting of 67 HCC, 111 liver cirrhotic, and 242 healthy plasma samples including 56 HBsAg+ samples.RESULTS: Based on methylation profiling of tissue samples, 2321 DMBs were identified, which were subsequently used to construct a cfDNA-based HCC screening model, achieved a sensitivity of 86% and specificity of 98% in the training cohort and a sensitivity of 84% and specificity of 96% in the independent validation cohort. This model obtained a sensitivity of 76% in 37 early-stage HCC (Barcelona clinical liver cancer [BCLC] stage 0-A) patients. The screening model can effectively discriminate HCC patients from non-HCC controls, including liver cirrhotic patients, asymptomatic HBsAg+ and healthy individuals, achieving an AUC of 0.957(95% CI 0.939-0.975), whereas serum α-fetoprotein (AFP) only achieved an AUC of 0.803 (95% CI 0.758-0.847). Besides detecting patients with early-stage HCC from non-HCC controls, this model showed high capacity for distinguishing early-stage HCC from a high risk population (AUC=0.934; 95% CI 0.905-0.963), also significantly outperforming AFP. Furthermore, our model also showed superior performance in distinguishing HCC with normal AFP (< 20ng ml-1) from high risk population (AUC=0.93; 95% CI 0.892-0.969).CONCLUSIONS: We have developed a sensitive blood-based non-invasive HCC screening model which can effectively distinguish early-stage HCC patients from high risk population and demonstrated its performance through an independent validation cohort.TRIAL REGISTRATION: The study was approved by the ethic committee of The Second Xiangya Hospital of Central South University (KYLL2018072) and Chongqing University Cancer Hospital (2019167). The study is registered at ClinicalTrials.gov(# NCT04383353 ). <<<

Probabilistic forecasting, i.e., estimating a time series’ future probability distribution given its past, is a key enabler for optimizing business processes. In retail businesses, for example, probabilistic demand forecasts are crucial for having the right inventory available at the right time and in the right place. This paper proposes DeepAR, a methodology for producing accurate probabilistic forecasts, based on training an autoregressive recurrent neural network model on a large number of related time series. We demonstrate how the application of deep learning techniques to forecasting can overcome many of the challenges that are faced by widely-used classical approaches to the problem. By means of extensive empirical evaluations on several real-world forecasting datasets, we show that our methodology produces more accurate forecasts than other state-of-the-art methods, while requiring minimal manual work. <<<

Simultaneous localization and mapping (SLAM) represents a fundamental problem for autonomous embodied systems, for which the hippocampal/entorhinal system (H/E-S) has been optimized over the course of evolution. We have developed a biologically-inspired SLAM architecture based on latent variable generative modeling within the Free Energy Principle and Active Inference (FEP-AI) framework, which affords flexible navigation and planning in mobile robots. We have primarily focused on attempting to reverse engineer H/E-S ‘design’ properties, but here we consider ways in which SLAM principles from robotics may help us better understand nervous systems and emergent minds. After reviewing LatentSLAM and notable features of this control architecture, we consider how the H/E-S may realize these functional properties not only for physical navigation, but also with respect to high-level cognition understood as generalized simultaneous localization and mapping (G-SLAM). We focus on loop-closure, graph-relaxation, and node duplication as particularly impactful architectural features, suggesting these computational phenomena may contribute to understanding cognitive insight (as proto-causal-inference), accommodation (as integration into existing schemas), and assimilation (as category formation). All these operations can similarly be describable in terms of structure/category learning on multiple levels of abstraction. However, here we adopt an ecological rationality perspective, framing H/E-S functions as orchestrating SLAM processes within both concrete and abstract hypothesis spaces. In this navigation/search process, adaptive cognitive equilibration between assimilation and accommodation involves balancing tradeoffs between exploration and exploitation; this dynamic equilibrium may be near optimally realized in FEP-AI, wherein control systems governed by expected free energy objective functions naturally balance model simplicity and accuracy. With respect to structure learning, such a balance would involve constructing models and categories that are neither too inclusive nor exclusive. We propose these (generalized) SLAM phenomena may represent some of the most impactful sources of variation in cognition both within and between individuals, suggesting that modulators of H/E-S functioning may potentially illuminate their adaptive significances as fundamental cybernetic control parameters. Finally, we discuss how understanding H/E-S contributions to G-SLAM may provide a unifying framework for high-level cognition and its potential realization in artificial intelligences. <<<

Forecasting is a common data science task that helps organizations with capacity planning, goal setting, and anomaly detection. Despite its importance, there are serious challenges associated with producing reliable and high quality forecasts — especially when there are a variety of time series and analysts with expertise in time series modeling are relatively rare. To address these challenges, we describe a practical approach to forecasting “at scale” that combines configurable models with analyst-in-the-loop performance analysis. We propose a modular regression model with interpretable parameters that can be intuitively adjusted by analysts with domain knowledge about the time series. We describe performance analyses to compare and evaluate forecasting procedures, and automatically flag forecasts for manual review and adjustment. Tools that help analysts to use their expertise most effectively enable reliable, practical forecasting of business time series. <<<