PaperPlayer biorxiv bioinformatics

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.10.13.337865v1?rss=1

Authors: Altschuler, S. J., Ji, W., Wu, L. F.

Abstract:

During development, neurons extend in search of synaptic partners. Precise control of axon extension velocity can therefore be crucial to ensuring proper circuit formation. How velocity is regulated - particularly by the extending axons themselves - remains poorly understood. Here, we investigate this question in the Drosophila visual system, where photoreceptors make precise connections with a specific set of synaptic partners that together create a circuit underpinning neural superposition (NSP). We used a combination of genetic perturbations and quantitative image analysis to investigate the influence of cell identity on growth cone velocity and subsequent spatial-temporal coincidence of presynaptic and postsynaptic neurons. Our study provides a case study of how cell autonomous properties of presynaptic axons play a pivotal role in controlling the dynamics of growing axons and determining the formation of a precise neuronal circuit.

Copy rights belong to original authors. Visit the link for more info

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.10.13.337147v1?rss=1

Authors: Nair, S., Shrikumar, A., Kundaje, A.

Abstract:

Deep learning models such as convolutional neural networks are able to accurately map biological sequences to associated functional readouts and properties by learning predictive de novo representations. In-silico saturation mutagenesis (ISM) is a popular feature attribution technique for inferring contributions of all characters in an input sequence to the model's predicted output. The main drawback of ISM is its runtime, as it involves multiple forward propagations of all possible mutations of each character in the input sequence through the trained model to predict the effects on the output. We present fastISM, an algorithm that speeds up ISM by a factor of over 10x for commonly used convolutional neural network architectures. fastISM is based on the observations that the majority of computation in ISM is spent in convolutional layers, and a single mutation only disrupts a limited region of intermediate layers, rendering most computation redundant. fastISM reduces the gap between backpropagation-based feature attribution methods and ISM. It far surpasses the runtime of backpropagation-based methods on multi-output architectures, making it feasible to run ISM on a large number of sequences. An easy-to-use Keras/TensorFlow 2 implementation of fastISM is available at https://github.com/kundajelab/fastISM, and a hands-on tutorial at https://colab.research.google.com/github/kundajelab/fastISM/blob/master/notebooks/colab/DeepSEA.ipynb.

Copy rights belong to original authors. Visit the link for more info

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.10.13.336826v1?rss=1

Authors: Xue, F., Tang, X., Kim, G., Aiello, A., Koenen, K., Galea, S., Wildman, D., Uddin, M., Qu, A.

Abstract:

Mediation analysis is widely used to understand mediating mechanisms of variables in causal inference. However, existing approaches do not consider heterogeneity in mediation effects. Mediators in different sub-populations could have opposite effects on the outcome, and could be difficult to identify under the homogeneous model framework. In this paper, we propose a new mediator selection method, which can identify sub-populations and select mediators in each sub-population for heterogeneous data simultaneously. We perform a multi-directional clustering analysis to determine sub-group mediators and the corresponding subjects. Specifically, to select mediators, we propose a new joint penalty which penalizes the effect of independent variable on a mediator and the effect of a mediator on the response jointly. The proposed algorithm is implemented through the convex-smooth gradient descent. Our numerical studies show that the proposed method outperforms the existing methods for heterogeneous data. We also apply the proposed mediation method to estimate mediation effects of DNA methylation variations in glucocorticoid receptor regulatory network genes for post-traumatic stress disorder (PTSD) among African-Americans. Based on data from the Detroit Neighborhood Health Study, we have found heterogeneous mediators, which are indeed associated with PTSD or traumatic experiences according to literature but were not selected by existing homogeneous mediation selection methods.

Copy rights belong to original authors. Visit the link for more info

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.10.12.335364v1?rss=1

Authors: Nyström-Persson, J. T., Keeble-Gagnere, G., Zawad, N.

Abstract:

The processing of k-mers (subsequences of length k) is at the foundation of many sequence processing algorithms in bioinformatics, including k-mer counting for genome size estimation, genome assembly, and taxonomic classification for metagenomics. Minimizers - ordered m-mers where m < k - are often used to group k-mers into bins as a first step in such processing. However, minimizers are known to generate bins of very different sizes, which can pose challenges for distributed and parallel processing, as well as generally increase memory requirements. Furthermore, although various minimizer orderings have been proposed, their practical value for improving tool efficiency has not yet been fully explored. Here we present a distributed k-mer counting tool based on Apache Spark, Discount, which we use to investigate the behaviour of various minimizer orderings in practice when applied to metagenomics data. Using this tool, we then introduce a new combination of frequency counted minimizers and universal k-mer hitting sets, the universal frequency ordering, which yields both evenly distributed binning and small bin sizes. We show that this new ordering allows Discount to perform distributed k-mer counting on a large dataset in as little as 1/8 of the memory of comparable approaches, making it the most efficient out-of-core distributed k-mer counting method available.

Copy rights belong to original authors. Visit the link for more info

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.10.12.336859v1?rss=1

Authors: Xu, J., Mcpartlon, M., Li, J.

Abstract:

We describe our latest study of the deep convolutional residual neural networks (ResNet) for protein structure prediction, including deeper and wider ResNets, the efficacy of different input features, and improved 3D model building methods. Our ResNet can predict correct folds (TMscore>0.5) for 26 out of 32 CASP13 FM (template-free-modeling) targets and L/5 long-range contacts for these targets with precision over 80%, a significant improvement over the CASP13 results. Although co-evolution analysis plays an important role in the most successful structure prediction methods, we show that when co-evolution is not used, our ResNet can still predict correct folds for 18 of the 32 CASP13 FM targets including several large ones. This marks a significant improvement over the top co-evolution-based, non-deep learning methods at CASP13, and other non-coevolution-based deep learning models, such as the popular recurrent geometric network (RGN). With only primary sequence, our ResNet can also predict correct folds for all 21 human-designed proteins we tested. In contrast, RGN predicts correct folds for only 3 human-designed proteins and zero CASP13 FM target. In addition, we find that ResNet may fare better for the human-designed proteins when trained without co-evolution information than with co-evolution. These results suggest that ResNet does not simply denoise co-evolution signals, but instead is able to learn important sequence-structure relationship from experimental structures. This has important implications on protein design and engineering especially when evolutionary information is not available.

Copy rights belong to original authors. Visit the link for more info

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.10.12.336784v1?rss=1

Authors: Hendley, A. M., Rao, A. A., Leonhardt, L., Ashe, S., Smith, J. A., Giacometti, S., Peng, X. L., Jiang, H., Berrios, D., Pawlak, M., Li, L. Y., Lee, J., Collisson, E. A., Anderson, M., Fragiadakis, G. K., Yeh, J. J., Chun, J. Y., Kim, G. E., Weaver, V. M., Hebrok, M.

Abstract:

Lineage tracing using genetically engineered mouse models is an essential tool for investigating cell-fate decisions of progenitor cells and biology of mature cell types, with relevance to physiology and disease progression. To study disease development, an inventory of an organ's cell types and understanding of physiologic function is paramount. Here, we performed single-cell RNA sequencing to examine heterogeneity of murine pancreatic duct cells, pancreatobiliary cells, and intrapancreatic bile duct cells. We isolated duct cells within the murine pancreas using a Dolichos biflorus agglutinin (DBA) lectin sorting strategy that labels all pancreatic duct cell types. Our data suggested the substructure of murine pancreatic duct cells is compartmentalized into three subpopulations. We describe an epithelial-mesenchymal transitory axis in our three pancreatic duct subpopulations and identify SPP1 as a regulator of this fate decision as well as human duct cell de-differentiation. Our results further identify functional heterogeneity within pancreatic duct subpopulations by elucidating a role for Geminin in accumulation of DNA damage in the setting of chronic pancreatitis. Our findings implicate diverse functional roles for subpopulations of pancreatic duct cells in maintenance of duct cell identity and disease progression and establish a comprehensive road map of murine pancreatic duct cell, pancreatobiliary cell, and intrapancreatic bile duct cell homeostasis.

Copy rights belong to original authors. Visit the link for more info

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.10.12.336388v1?rss=1

Authors: NAVARRO, R. C., Ligia Scott, A., Allison Philot, E., Atencio, L., Fernandez Ponce, C., Aroca Martinez, G., Cadena Bonfanti, A., Gomez Escorcia, l., Navarro Quiroz, E.

Abstract:

Abstract: Background: The human dopamine transporter is the main regulator of dopamine tone and an intricate network exists to regulate the expression, conformation, and kinetics of the hDAT. hDAT dysfunction is directly related to Parkinson's disease. The objective of this work is to evaluate the differential gene expression in the interactome of the Dopamine transporter in the context of Parkinson's disease. Methods:To do this, we evaluated hDAT interaction data in string-db, mint.bio, IntAct, reactome, hprd and BioGRID, subsequently, data was obtained from the differential gene expression of mRNA and miRNAs for this hDAT interactome in the context of PD. Results: The analysis of the differential expression changes of genes of the hDAT interactome in tissues of patients with PD compared with tissues of individuals without PD, allowed to identify an expression pattern of 32 components of the hDAT interactome, of which 31 presented a negative change proportion in PD. Conclusions: We found a total of 90 miRNAs that could regulate the expression of 27 components of the hDAT interactome, at the same time, 39 components of the hDAT interactome may participate in 40 metabolic pathways. Together, these findings show a systematic effect on the hDAT-mediated dopamine internalization process in patients with Parkinson's, which would contribute to a greater susceptibility to neuronal oxidative stress in PD patients.

Copy rights belong to original authors. Visit the link for more info

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.10.12.336404v1?rss=1

Authors: Candia, J.

Abstract:

Summary: mutSigMapper aims to resolve a critical shortcoming of existing software for mutational signature analysis, namely that of finding parsimonious and biologically plausible exposures. By implementing a shot-noise-based model to generate spectral ensembles, this package addresses this gap and provides a quantitative, non-parametric assessment of statistical significance for the association between mutational signatures and observed spectra. Availability and implementation: The mutSigMapper R package is available under GPLv3 license at https://github.com/juliancandia/mutSigMapper. Its documentation provides additional details and demonstrates applications to biological datasets.

Copy rights belong to original authors. Visit the link for more info

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.10.12.336271v1?rss=1

Authors: Zhao, Z., Woloszynek, S., Agbavor, F., Mell, J., Sokhansanj, B. A., Rosen, G.

Abstract:

Recurrent neural networks (RNNs) with memory (e.g. LSTMs) and attention mechanisms are widely used in natural language processing because they can capture short and long term sequential information for diverse tasks. We propose an integrated deep learning model for microbial DNA sequence data, which exploits convolutional networks, recurrent neural networks, and attention mechanisms to perform sample-associated attribute prediction---phenotype prediction---and extract interesting features, such as informative taxa and predictive k-mer context. In this paper, we develop this novel deep learning approach and evaluate its application to amplicon sequences. We focus on typically short DNA reads of 16s ribosomal RNA (rRNA) marker genes, which identify the heterogeneity of a microbial community sample. Our deep learning approach enables sample-level attribute and taxonomic prediction, with the aim of aiding biological research and supporting medical diagnosis. We demonstrate that our implementation of a novel attention-based deep network architecture, Read2Pheno, achieves read-level phenotypic prediction and, in turn, that aggregating read-level information can robustly predict microbial community properties, host phenotype, and taxonomic classification, with performance comparable to conventional approaches. Most importantly, as a further result of the training process, the network architecture will encode sequences (reads) into dense, meaningful representations: learned embedded vectors output on the intermediate layer of the network model, which can provide biological insight when visualized. Finally, we demonstrate that a model with an attention layer can automatically identify informative regions in sequences/reads which are particularly informative for classification tasks. An implementation of the attention-based deep learning network is available at https://github.com/EESI/sequence_attention.

Copy rights belong to original authors. Visit the link for more info

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.10.12.318808v1?rss=1

Authors: R Melo Costa, V., Pfeuffer, J., Louloupi, A., V Orom, U. A., Piro, R. M.

Abstract:

Background: Introns are generally removed from primary transcripts to form mature RNA molecules in a post-transcriptional process called splicing. An efficient splicing of primary transcripts is an essential step in gene expression and its misregulation is related to numerous human diseases. Thus, to better understand the dynamics of this process and the perturbations that might be caused by aberrant transcript processing it is important to quantify splicing efficiency. Results: Here, we introduce SPLICE-q, a fast and user-friendly Python tool for genome-wide SPLICing Efficiency quantification. It supports studies focusing on the implications of splicing efficiency in transcript processing dynamics. SPLICE-q uses aligned reads from strand-specific RNA-seq to quantify splicing efficiency for each intron individually and allows the user to select different levels of restrictiveness concerning the introns' overlap with other genomic elements such as exons of other genes. We applied SPLICE-q to globally assess the dynamics of intron excision in yeast and human nascent RNA-seq. We also show its application using total RNA-seq from a patient-matched prostate cancer sample. Conclusions: Our analyses illustrate that SPLICE-q is suitable to detect a progressive increase of splicing efficiency throughout a time course of nascent RNA-seq and it might be useful when it comes to understanding cancer progression beyond mere gene expression levels. SPLICE-q is available at: github.com/vrmelo/SPLICE-q .

Copy rights belong to original authors. Visit the link for more info