PaperPlayer biorxiv bioinformatics

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.08.06.239988v1?rss=1

Authors: Liu, Y., Lim, H., Xie, L.

Abstract:

Motivation Drug discovery is time-consuming and costly. Machine learning, especially deep learning, shows a great potential in accelerating the drug discovery process and reducing the cost. A big challenge in developing robust and generalizable deep learning models for drug design is the lack of data with high quality and balanced labels. To address this challenge, we developed a self-training method PLANS that exploits millions of unlabeled chemical compounds as well as partially labeled pharmacological data to improve the generalizability of neural network models. Result We evaluated the self-training with PLANS for Cytochrome P450 binding activity prediction task, and proved that our method could significantly improve the performance of the neural network model with a large margin. Compared with the baseline model, the PLANS-trained neural network model improved accuracy, precision, recall, and F1 score by 13.4%, 12.5%, 8.3%, and 10.3%, respectively. The self-training with PLANS is model agnostic, and can be applied to any deep learning architectures. Thus, PLANS provides a general solution to utilize unlabeled and partially labeled data to improve the predictive modeling for drug discovery.

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

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.08.06.239244v1?rss=1

Authors: Huber, F., Verhoeven, S., Meijer, C., Spreeuw, H., Villanueva Castilla, E. M., Geng, C., van der Hooft, J. J., Rogers, S., Belloum, A., Diblen, F., Spaaks, J. H.

Abstract:

Mass spectrometry data is at the heart of numerable applications in the biomedical and life sciences. With growing use of high throughput techniques researchers need to analyse larger and more complex datasets. In particular through joint effort in the research community, fragmentation mass spectrometry datasets are growing in size and number. Matchms is an open-access Python package to import, process, clean, and compare mass spectrometry data (MS/MS). It allows to implement and run an easy-to-follow, easy-to-reproduce workflow from raw mass spectra to pre- and post-processed spectral data.

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

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.08.06.239905v1?rss=1

Authors: Kugler, E. C., Frost, J., Silva, V., Plant, K. C., Chhabria, K., Chico, T. J., Armitage, P.

Abstract:

Zebrafish transgenic lines and light sheet fluorescence microscopy allow in-depth insights into vascular development in vivo and 3D. However, robust quantification of the zebrafish cerebral vasculature in 3D remains a challenge, and would be essential to describe the vascular architecture. Here, we report an image analysis pipeline that allows 3D quantification of the total or regional zebrafish brain vasculature. This is achieved by landmark- or object-based inter-sample registration and extraction of quantitative parameters including vascular volume, surface area, density, branching points, length, radius, and complexity. Application of our analysis pipeline to a range of sixteen genetic or pharmacological manipulations shows that our quantification approach is robust, allows extraction of biologically relevant information, and provides novel insights into vascular biology. To allow dissemination, the code for quantification, a graphical user interface, and workflow documentation are provided. Together, we present the first 3D quantification approach to assess the whole 3D cerebrovascular architecture in zebrafish.

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

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.08.05.238295v1?rss=1

Authors: Mehrotra, S., Bronstein, R., Navarro-Gomez, D., Segre, A. V., Pierce, E. A.

Abstract:

High-throughput transcriptome sequencing has become a powerful tool in the study of human diseases. Identification of causal mechanisms may entail analysis of differential gene expression (DGE), differential transcript/isoform expression (DTE) and identification, classification and quantification of alternative splicing (AS) and/or detection of novel AS events. For such a global transcriptome profiling execution of multi-level data analysis methodologies is required. Each level presents its own unique challenges and the questions about their performance remains. In this work we present results from systematic and consistent assessing and comparing a number of widely used methods for detecting DGE, DTE and AS using internal control "spike-in" sequences (Sequins) in RNA-seq data. We demonstrated that inclusion of internal controls in RNA-seq experiments allows accurate determination of lower bounds detection levels, and better assessment of DGE, DTE and AS accuracy and sensitivity. Tools for RNA-seq read alignment and detection of DGE performed reasonably. More efforts are needed to improve specificity and sensitivity of DTE and AS detection. Low expression of isoforms accompanied with sequencing depth does impact sensitivity and specificity of DTE and AS tools.

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

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.08.05.238519v1?rss=1

Authors: Chereda, H., Bleckmann, A., Menck, K., Perera-Bel, J., Stegmaier, P., Auer, F., Kramer, F., Leha, A., Beissbarth, T.

Abstract:

Motivation: Contemporary deep learning approaches show cutting-edge performance in a variety of complex prediction tasks. Nonetheless, the application of deep learning in healthcare remains limited since deep learning methods are often considered as non-interpretable black-box models. Layer-wise Relevance Propagation (LRP) is a technique to explain decisions of deep learning methods. It is widely used to interpret Convolutional Neural Networks (CNNs) applied on image data. Recently, CNNs started to extend towards non-euclidean domains like graphs. Molecular networks are commonly represented as graphs detailing interactions between molecules. Gene expression data can be assigned to the vertices of these graphs. In other words, gene expression data can be structured by utilizing molecular network information as prior knowledge. Graph-CNNs can be applied to structured gene expression data, for example, to predict metastatic events in breast cancer. Therefore, there is a need for explanations showing which part of a molecular network is relevant for predicting an event, e.g. distant metastasis in cancer, for each individual patient. Results: We extended the procedure of LRP to make it available for Graph-CNN and tested its applicability on a large breast cancer dataset. We present Graph Layer-wise Relevance Propagation (GLRP) as a new method to explain the decisions made by Graph-CNNs. We demonstrate a sanity check of the developed GLRP on a hand-written digits dataset, and then applied the method on gene expression data. We show that GLRP provides patient-specific molecular subnetworks that largely agree with clinical knowledge and identify common as well as novel, and potentially druggable, drivers of tumor progression. As a result this method could be potentially highly useful on interpreting classification results on the individual patient level, as for example in precision medicine approaches or a molecular tumor board.

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

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.08.04.232371v1?rss=1

Authors: HarshaRani, G. V., Moza, S., Ramakrishnan, N., Bhalla, U. S.

Abstract:

Bistable biochemical switches are key motifs in cellular state decisions and long-term storage of cellular 'memory'. There are a few known biological switches that have been well characterized, however these examples are insufficient for systematic surveys of properties of these important systems. Here we present a resource of all possible bistable biochemical reaction networks with up to 6 reactions between 3 molecules, and 3 reactions between 4 molecules. Over 35,000 reaction topologies were constructed by identifying unique combinations of reactions between a fixed number of molecules. Then, these topologies were populated with rates within a biologically realistic range. The Searchable Web Interface for Topologies of CHEmical Switches (SWITCHES, https://switches.ncbs.res.in) provides a bistability and parameter analysis of over 7 million models from this systematic survey of chemical reaction space. This database will be useful for theoreticians interested in analyzing stability in chemical systems and also experimentalists for creating robust synthetic biological switches.

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

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.08.02.232645v1?rss=1

Authors: Moolamalla, S. T. R., Chauhan, R., Priyakumar, U. D., Vinod, P. K.

Abstract:

Understanding the pathogenesis of SARS-CoV-2 is important for developing effective treatment strategies. Viruses hijack the host metabolism to redirect the resources for their replication and survival. How SARS-CoV-2 influences the host metabolism is still unclear. In this study, we analyzed transcriptomic data obtained from different human respiratory cell lines and patient samples (Swab, PBMC, lung biopsy, BALF) to understand the metabolic alterations in response to SARS-CoV-2 infection. For this purpose, the expression pattern of metabolic genes in the human genome-scale metabolic network model Recon3D was explored. We identified metabolic genes and pathways and reporter metabolites under each SARS-CoV-2-infected condition and compared them to identify common and unique changes in the metabolism. Our analysis revealed host-dependent dysregulation of glycolysis, mitochondrial metabolism, amino acid metabolism, glutathione metabolism, polyamine synthesis, and lipid metabolism. We observed different metabolic changes that are pro- and antiviral in nature. We generated hypotheses on how antiviral metabolism can be targeted/enhanced for reducing viral titers. These warrant further exploration with more samples and in vitro studies to test predictions.

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

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.08.04.236646v1?rss=1

Authors: Xiao, W., Zhang, X., Xiao, W.

Abstract:

Motivation: Essential genes are necessary to the survival or reproduction of a living organism. The prediction and analysis of gene essentiality can advance our understanding to basic life and human diseases, and further boost the development of new drugs. Wet lab methods for identifying essential genes are often costly, time consuming, and laborious. As a complement, computational methods have been proposed to predict essential genes by integrating multiple biological data sources. Most of these methods are evaluated on model organisms. However, prediction methods for human essential genes are still limited and the relationship between human gene essentiality and different biological information still needs to be explored. In addition, exploring suitable deep learning techniques to overcome the limitations of traditional machine learning methods and improve the prediction accuracy is also important and interesting. Results: We propose a deep learning based method, DeepSF, to predict human essential genes. DeepSF integrates four types of features, that is, sequence features, features from gene ontology, features from protein complex, and network features. Sequence features are derived from DNA and protein sequence for each gene. 100 GO terms from cellular component are used to form a feature vector for each gene, in which each component captures the relationship between a gene and a GO term. Network features are learned from protein-protein interaction (PPI) network using a deep learning based network embedding method. The features derived from protein complexes capture the relationships between a gene or a gene's direct neighbors from PPI network and protein complexes. The four types of features are integrated together to train a multilayer neural network. The experimental results of 10-fold cross validation show that DeepSF can accurately predict human gene essentiality with an average performance of AUC about 94.35%, the area under precision-recall curve (auPRC) about 91.28%, the accuracy about 91.35%, and the F1 measure about 77.79%. In addition, the comparison results show that DeepSF significantly outperforms several widely used traditional machine learning models (SVM, Random Forest, and Adaboost), and performs slightly better than a recent deep learning model (DeepHE). Conclusions: We have demonstrated that the proposed method, DeepSF, is effective for predicting human essential genes. Deep learning techniques are promising at both feature learning and classification levels for the task of essential gene prediction.

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

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.08.04.235168v1?rss=1

Authors: Danaher, P., Kim, Y., Nelson, B., Griswold, M., Yang, Z., Piazza, E., Beechem, J. M.

Abstract:

We introduce SpatialDecon, an algorithm for quantifying cell populations defined by single cell RNA sequencing within the regions of spatially-resolved gene expression studies. It obtains cell abundance estimates that are spatially-resolved, granular, and paired with highly multiplexed gene expression data. SpatialDecon incorporates several advancements in the field of gene expression deconvolution. We propose an algorithm based in log-normal regression, attaining sometimes dramatic performance improvements over classical least-squares methods. We compile cell profile matrices for 27 tissue types. We identify genes whose minimal expression by cancer cells makes them suitable for immune deconvolution in tumors. And we provide a lung tumor dataset for benchmarking immune deconvolution methods. In a lung tumor GeoMx DSP experiment, we observe a spatially heterogeneous immune response in intricate detail and identify 7 distinct phenotypes of the localized immune response. We then demonstrate how cell abundance estimates give crucial context for interpreting gene expression results.

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

Link to bioRxiv paper:

http://biorxiv.org/cgi/content/short/2020.08.03.232843v1?rss=1

Authors: Kirchner, T., Riege, K., Hoffmann, S.

Abstract:

Arguably, one of the most important tools for computer science is the Linux shell. Processing steps carried out there are critical for many analyses and software development projects. However, manual documentation of the work is time-consuming and error-prone. To remedy this problem, shournal tightly integrates with the shell and automatically records all shell commands along with associated file events. For any file, shournal allows the reconstruction of the command history and is able to create detailed reports for whole project directories. shournal is based on the fanotify API and mount namespaces and allows the efficient monitoring of entire process trees.

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