Sample multiplexing facilitates scRNA-seq by reducing costs and artifacts such as cell doublets. However, universal and scalable sample barcoding strategies have not been described. We therefore developed MULTI-seq: multiplexing using lipid-tagged indices for single-cell and single-nucleus RNA sequencing. MULTI-seq reagents can barcode any cell type or nucleus from any species with an accessible plasma membrane. The method involves minimal sample *

Summary The exocrine pancreas, consisting of ducts and acini, is the site of origin of pancreatitis and pancreatic ductal adenocarcinoma (PDAC). Our understanding of the genesis and progression of human pancreatic diseases, including PDAC, is limited because of challenges in maintaining human acinar and ductal cells in culture. Here we report induction of human pluripotent stem cells toward pancreatic ductal and acinar organoids that recapitulate properties of the neonatal exocrine pancreas. Expression of the PDAC-associated oncogene GNAS R201C induces cystic growth more effectively in ductal than acinar organoids, whereas KRAS G12D is more effective in modeling cancer in vivo when expressed in acinar compared with ductal organoids. KRAS G12D , but not GNAS R201C , induces acinar-to-ductal metaplasia-like changes in culture and in vivo . We develop a renewable source of ductal and acinar organoids for modeling exocrine development and diseases and demonstrate lineage tropism and plasticity for oncogene action in the human pancreas.

Detecting multiplets in single nucleus (sn)ATAC-seq data is challenging due to data sparsity and limited dynamic range. AMULET (ATAC-seq MULtiplet Estimation Tool) enumerates regions with greater than two uniquely aligned reads across the genome to effectively detect multiplets. We evaluate the method by generating snATAC-seq data in the human blood and pancreatic islet samples. AMULET has high precision, estimated via donor-based multiplexing, and high recall, estimated via simulated multiplets, compared to alternatives and identifies multiplets most effectively when a certain read depth of 25K median valid reads per nucleus is achieved.

Abstract-In this paper, a homography-based approach for determining the ground plane using image pairs is presented. Our approach is unique in that it uses a Modified Expectation Maximization algorithm to cluster pixels on images as belonging to one of two possible classes: ground and non-ground pixels. This classification is very useful in mobile robot navigation because, by segmenting out the ground plane, we are left with all possible objects in the scene, which can then be used to implement many mobile robot navigation algorithms such as obstacle avoidance, path planning, target following, landmark detection, etc. Specifically, we demonstrate the usefulness and robustness of our approach by applying it to a target following algorithm. As the results section shows, the proposed algorithm for ground plane detection achieves an almost perfect detection rate (over 99%) despite the relatively higher number of errors in pixel correspondence from the feature matching algorithm used: SIFT.

19We describe MULTI-seq: A rapid, modular, and universal scRNA-seq sample multiplexing 20 strategy using lipid-tagged indices. MULTI-seq reagents can barcode any cell type from any 21 species with an accessible plasma membrane. The method is compatible with enzymatic tissue 22 thus limited to quantifying 10s-100s of single-cell transcriptomes at a time (Tang et al., 2009; 38 Ramsköld et al., 2012; Hashimony et al., 2012). Today, the advent and commercialization of 39 microwell (Gierahn et al., 2017), split-pool barcoding (Rosenberg et al., 2018), and droplet-40 microfluidics (Macosko et al., 2015;Klein et al., 2015;Zheng et al., 2017) methods has enabled 41 the routine transcriptional analysis of 10 3 -10 5 cells in parallel. The essential insight enabling 42 these approaches is identical -pools of transcripts are linked to their cell-of-origin via DNA 43 barcodes introduced during reverse transcription and/or ligation. This enormous increase in cell 44 throughput enabled by these methods has catalyzed efforts to catalog the composition of whole 45 organs (The Tabula Muris Consortium et al., 2018) and even entire organisms (Cao et al., 2017; 46 Han et al., 2018). Indeed, ambitious efforts are now underway to create a cell-type atlas for the 47 human body using the latest scRNA-seq techniques (Regev et al., 2017). However, much as 48 research priorities shifted away from describing DNA sequences to functional genomics 49 following the culmination of the Human Genome Project (Lander et al., 2001; ENCODE Project 50 Consortium, 2012), the single-cell genomics field will soon expand beyond descriptive analyses 51 of cell types to mechanistically characterizing how these diverse cell populations interact through 52 space and time to regulate development, homeostasis, and disease. 53In order to utilize single-cell sequencing technologies to reveal mechanistic insights into 54 complex multicellular biology, the enormous throughput of scRNA-seq methods must be 55 redirected towards hypothesis testing. This requires integrating dynamical information, many 56 experimental perturbations, and multiple replicates in order to draw strong conclusions. While 57 existing methods are optimally configured to assay many thousands of cells, library preparation 58 practices and the physical constraints of current commercially-available microfluidic devices 59 (e.g., the Fluidigm C1 and 10X Genomics Single-Cell V2 systems) limit analyses to sets of 8 or 60 fewer conditions in a typical scRNA-seq experiment. Experiments that attempt to compare large 61 numbers of samples across multiple single-cell sequencing runs frequently suffer from batch 62 effects (Stegle et al., 2015;Haghverdi et al., 2018). Furthermore, at current prices, the reagent 63 and sequencing costs associated with analyzing large sample numbers is outside the means of 64 typical research groups. One approach to circumvent these challenges would be to sequence 65 large numbers of cells from diverse samples, but with relatively fewer cells from each sample. 66Enco...