Biomedical researchers are generating high-throughput, high-dimensional single-cell 5 data at a staggering rate. As costs of data generation decrease, experimental design is mov-6 ing towards measurement of many different single-cell samples in the same dataset. These 7 samples can correspond to different patients, conditions, or treatments. While scalability of 8 methods to datasets of these sizes is a challenge on its own, dealing with large-scale exper-9 imental design presents a whole new set of problems, including batch effects and sample 10 1 .

Handling the vast amounts of single-cell RNA-sequencing and CyTOF data, which are now being generated in patient cohorts, presents a computational challenge due to the noise, complexity, sparsity and batch effects present. Here, we propose a unified deep neural network-based approach to automatically process and extract structure from these massive datasets. Our unsupervised architecture, called SAUCIE (Sparse Autoencoder for Unsupervised Clustering, Imputation, and Embedding), simultaneously performs several key tasks for single-cell data analysis including 1) clustering, 2) batch correction, 3) visualization, and 4) denoising/imputation. SAUCIE is trained to recreate its own input after reducing its dimensionality in a 2-D embedding layer which can be used to visualize the data. Additionally, it uses two novel regularizations: (1) an information dimension regularization to penalize entropy as computed on normalized activation values of the layer, and thereby encourage binary-like encodings that are amenable to clustering and (2) a Maximal Mean Discrepancy penalty to correct batch effects. Thus SAUCIE has a single architecture that denoises, batch-corrects, visualizes and clusters data using a unified 1 . CC-BY 4.0 International license peer-reviewed) is the author/funder. It is made available under a

Interest in image-to-image translation has grown substantially in recent years with the success of unsupervised models based on the cycle-consistency assumption. The achievements of these models have been limited to a particular subset of domains where this assumption yields good results, namely homogeneous domains that are characterized by style or texture differences. We tackle the challenging problem of image-to-image translation where the domains are defined by high-level shapes and contexts, as well as including significant clutter and heterogeneity. For this purpose, we introduce a novel GAN based on preserving intra-domain vector transformations in a latent space learned by a siamese network. The traditional GAN system introduced a discriminator network to guide the generator into generating images in the target domain. To this two-network system we add a third: a siamese network that guides the generator so that each original image shares semantics with its generated version. With this new threenetwork system, we no longer need to constrain the generators with the ubiquitous cycle-consistency restraint. As a result, the generators can learn mappings between more complex domains that differ from each other by large differences -not just style or texture).

Type 1 diabetes (T1D) is most likely caused by killing of β cells by autoreactive CD8 T cells. Methods to isolate and identify these cells are limited by their low frequency in the peripheral blood. We analyzed CD8 T cells, reactive with diabetes Ags, with T cell libraries and further characterized their phenotype by CyTOF using class I MHC tetramers. In the libraries, the frequency of islet Ag-specific CD45ROIFN-γCD8 T cells was higher in patients with T1D compared with healthy control subjects. Ag-specific cells from the libraries of patients with T1D were reactive with ZnT8, whereas those from healthy control recognized ZnT8 and other Ags. ZnT8-reactive CD8 cells expressed an activation phenotype in T1D patients. We found TCR sequences that were used in multiple library wells from patients with T1D, but these sequences were private and not shared between individuals. These sequences could identify the Ag-specific T cells on a repeated draw, ex vivo in the IFN-γ CD8 T cell subset. We conclude that CD8 T cell libraries can identify Ag-specific T cells in patients with T1D. The T cell clonotypes can be tracked in vivo with identification of the TCR gene sequences.

The genus Flavivirus contains many mosquito-borne human pathogens of global epidemiological importance such as dengue virus, West Nile virus, and Zika virus, which has recently emerged at epidemic levels. Infections with these viruses result in divergent clinical outcomes ranging from asymptomatic to fatal. Myriad factors influence infection severity including exposure, immune status and pathogen/host genetics. Furthermore, pre-existing infection may skew immune pathways or divert immune resources. We profiled immune cells from dengue virus-infected individuals by multiparameter mass cytometry (CyTOF) to define functional status. Elevations in IFNβ were noted in acute patients across the majority of cell types and were statistically elevated in 31 of 36 cell subsets. We quantified response to in vitro (re)infection with dengue or Zika viruses and detected a striking pattern of upregulation of responses to Zika infection by innate cell types which was not noted in response to dengue virus. Significance was discovered by statistical analysis as well as a neural network-based clustering approach which identified unusual cell subsets overlooked by conventional manual gating. Of public health importance, patient cells showed significant enrichment of innate cell responses to Zika virus indicating an intact and robust anti-Zika response despite the concurrent dengue infection.

Conclusion: There was a significant decline in memory in PCI group. Further investigation to assess its impact on long-term follow-up is in progress.

Recurrent neural networks have achieved remarkable success at generating sequences with complex structures, thanks to advances that include richer embeddings of input and cures for vanishing gradients. Trained only on sequences from a known grammar, though, they can still struggle to learn rules and constraints of the grammar. Neural Attribute Machines (NAMs) are equipped with a logical machine that represents the underlying grammar, which is used to teach the constraints to the neural machine by (i) augmenting the input sequence, and (ii) optimizing a custom loss function. Unlike traditional RNNs, NAMs are exposed to the grammar, as well as samples from the language of the grammar. During generation, NAMs make significantly fewer violations of the constraints of the underlying grammar than RNNs trained only on samples from the language of the grammar.

While generative models such as GANs have been successful at mapping from noise to specific distributions of data, or more generally from one distribution of data to another, they cannot isolate the transformation that is occurring and apply it to a new distribution not seen in training. Thus, they memorize the domain of the transformation, and cannot generalize the transformation out of sample. To address this, we propose a new neural network called a Neuron Transformation Network (NTNet) that isolates the signal representing the transformation itself from the other signals representing internal distribution variation. This signal can then be removed from a new dataset distributed differently from the original one trained on. We demonstrate the effectiveness of our NTNet on more than a dozen synthetic and biomedical single-cell RNA sequencing datasets, where the NTNet is able to learn the data transformation performed by genetic and drug perturbations on one sample of cells and successfully apply it to another sample of cells to predict treatment outcome.