Biocomputing 2018 - Proceedings Of The Pacific Symposium.

• Author: Altman, Russ B.
• TeilnehmendeR: Dunker, A Keith.; Hunter, Lawrence.; Ritchie, Marylyn D.; Murray, Tiffany A.; Klein, Teri E.
• Place / Publishing House: Singapore : : World Scientific Publishing Company,, 2017.; Ã2018.
• Year of Publication: 2017
• Edition: 1st ed.
• Language: English
• Physical Description: 1 online resource (649 pages)

Table of Contents

• Intro
• Preface
• APPLICATIONS OF GENETICS, GENOMICS AND BIOINFORMATICS IN DRUG DISCOVERY
• Session introduction
• 1. Introduction
• 2. Session Contributions
• 2.1. Drug mechanisms of action and drug combinations
• 2.2. Drug metabolism and in silico drug screening
• 2.3. Disease genes and pathways
• 3. Acknowledgments
• References
• Characterization of drug-induced splicing complexity in prostate cancer cell line using long read technology
• Introduction
• Results
• Discussion
• Methods
• Supplementary
• Acknowledgements
• References
• Prediction of protein-ligand interactions from paired protein sequence motifs and ligand substructures
• 1. Introduction
• 1.1. Decreasing returns in drug discovery pipelines
• 1.2. Existing methods for prediction of protein-ligand interactions
• 2. Methods
• 2.1. Data set
• 2.2. Protein Featurization
• 2.3. Ligand Featurization
• 2.4. Boosting Model
• 2.5. Cross Validation Approaches
• 3. Results
• 3.1. Model Performance
• 3.2. Most predictive motif features
• 3.3. Known positive examples
• 3.3.1. Uricase - Uric acid
• 3.3.2. Chloramphenicol O-acetyltransferase - Chloramphenicol
• 3.3.3. Transthyretin -T4
• 3.4. Interpreting ADT Paths
• 3.4.1. Path lengths
• 3.4.2. Protein kinase C - Phosphatidylserine
• 4. Discussion
• Acknowledgments
• References
• Cell-specific prediction and application of drug-induced gene expression profiles
• 1. Introduction
• 2. Methods
• 2.1. Notation and terminology
• 2.2. Data processing
• 2.3. The Drug Neighbor Profile Prediction algorithm
• 2.4. The Fast, Low-Rank Tensor Completion algorithm
• 2.5. Baseline averaging schemes
• 2.6. Cross-validation for predicting gene expression profiles
• 2.7. Predicting drug targets and ATC codes
• 3. Results
• 3.1. Overall accuracy
• 3.2. Tradeoffs in accuracy across drug-cell space.
• 3.3. Effects of varying observation density
• 3.4. Accuracy of differentially expressed genes
• 3.5. Analysis of cell-specificity
• 3.6. Utility of completed data for downstream prediction of drug properties
• 4. Discussion
• Supplementary Information
• Funding
• References
• Large-scale integration of heterogeneous pharmacogenomic data for identifying drug mechanism of action
• 1. Introduction
• 2. Materials and Methods
• 2.1. Construction of heterogeneous drug-drug similarity networks
• 2.2. Integration of multi-omics data
• 2.3. Prediction of MoAs and drug targets
• 3. Results
• 3.1. Mania improves the quanti cation of drug-drug similarity
• 3.2. Mania achieves accurate prediction of drug MoAs and targets
• 3.3. Identification of functionally-enriched drug communities
• 3.4. Predictions of drugs for significantly mutated genes
• 4. Discussion
• References
• Chemical reaction vector embeddings: towards predicting drug metabolism in the human gut microbiome
• 1. Introduction
• 2. Methods
• 2.1. Data sources and processing
• 2.2. Constructing molecular vector space
• 2.3. Characterizing vector spaces
• 2.3.1. Molecule-level Analysis
• 2.3.2. Reaction-Level Analysis
• 2.4. Querying drug-metabolite pairs against reaction vectors
• 3. Results
• 3.1. Molecule-level analysis
• 3.2. Reaction-level analysis
• 3.3. Querying reaction vectors against drug-metabolite pairs
• 4. Discussion
• 5. Conclusion
• 6. Acknowledgments
• References
• Loss-of-function of neuroplasticity-related genes confers risk for human neurodevelopmental disorders
• 1. Introduction
• 2. Methods
• 2.1 Neuroplasticity signatures
• 2.2 Hospital and biobank cohort
• 2.3 Variant annotation
• 2.4 Neurodevelopmental disease phenotyping
• 2.5 LOF gene and disease association analysis
• 3. Results
• 3.1 Identifying putative neuroplasticity genes.
• 3.2 LOF variants in putative plasticity genes confer risk for neurodevelopmental and nervous system related disorders
• 4. Discussion
• 5. Conclusions and Future Directions
• 6. Acknowledgments
• References
• Extracting a biologically relevant latent space from cancer transcriptomes with variational autoencoders
• 1. Introduction
• 2. Methods
• 2.1. Model Summary
• 2.2. Model Implementation
• 2.3. Parameter Selection
• 2.4. Input Data
• 2.5. Interpretation of Gene Weights
• 2.6. The Latent Space of Ovarian Cancer Subtypes
• 2.7. Enabling Exploration through Visualization
• 3. Results
• 3.1. Tumors were encoded in a lower dimensional space
• 3.2. Features represent biological signal
• 3.3. Interpolating the lower dimensional manifold of HGSC subtypes
• 4. Conclusion
• 5. Reproducibility
• Acknowledgments
• References
• Diffusion mapping of drug targets on disease signaling network elements reveals drug combination strategies
• 1. Introduction
• 2. Methods
• 3. Results
• 4. Discussion
• References
• CHALLENGES OF PATTERN RECOGNITION IN BIOMEDICAL DATA
• Session introduction
• 1. Introduction
• 2. Session Contributions
• 2.1 Network-based approaches
• 2.2 Machine learning approaches
• 2.2 Application of methods to identify patterns in EHR data
• 2.3 Applications in transcriptome and next-generation sequencing data
• 3. References
• Large-scale analysis of disease pathways in the human interactome
• 1. Introduction
• 2. Background and related work
• 3. Data
• 4. Connectivity of disease proteins in the PPI network
• 4.1. Proximity of disease proteins in the PPI network
• 4.2. Connections between PPI network structure and disease protein discovery
• 5. Higher-order connectivity of disease proteins in the PPI network
• 6. Prediction of disease proteins using higher-order PPI network structure
• 7. Conclusion.
• Acknowledgments
• References
• Mapping patient trajectories using longitudinal extraction and deep learning in the MIMIC-III Critical Care Database
• 1. Introduction
• 2. Methods
• 2.1. Source Code and Analysis Availability
• 2.2. Care Event Extraction
• 2.3. Unsupervised learning to learn embeddings of extracted Care Events
• 2.4. Predicting Survival Using Care Events
• 3. Results
• 3.1. Treatment and Outcome Comparison
• 3.2. Unsupervised modeling of patient care events
• 3.3. Supervised prediction of patient survival
• 4. Discussion and Conclusions
• 5. Acknowledgments
• References
• OWL-NETS: Transforming OWL representations for improved network inference
• 1. Introduction
• 2. Methods
• 2.1. Biomedical Use Cases
• 2.2. Link Prediction Procedures
• 2.2.1. Evaluation of Link Prediction Algorithm Performance
• 2.2.2. Evaluation of Inferred Edges
• 3. Results
• 3.1. Comparison of Network Properties
• 3.2. Link Prediction Algorithm Performance
• 3.2.1. Inferred Edges
• 4. Discussion
• 5. Conclusions
• 6. Acknowledgments
• 7. Funding
• References
• Automated disease cohort selection using word embeddings from Electronic Health Records
• 1. Introduction
• 2. Methods and Materials
• 2.1. Research Cohort and Resource
• 2.2. Disease Phenotyping Algorithms
• 2.3. Phenotype and Patient Embedding
• 2.4. Evaluation Design
• 3. Results
• 3.1. Evaluating Performance of Embeddings
• 4. Discussion
• 4.1. Limitations and Future Directions
• 5. Acknowledgments
• References
• Functional network community detection can disaggregate and filter multiple underlying pathways in enrichment analyses
• 1. Introduction
• 2. Methods
• 2.1. General Approach
• 2.2. Control Arm
• 2.3. Experimental Arm
• 3. Results and Discussion
• 3.1. Simulation Study
• 3.2. HGSC Results
• 4. Conclusion
• 5. Acknowledgments.
• 6. Supplementary Material
• References
• An ultra-fast and scalable quantification pipeline for transposable elements from next generation sequencing data
• 1. Introduction
• 2. Methods
• 2.1. Transposable Element Library Preparation
• 2.2. Salmon quanti cation algorithm
• 2.3. Statistical tests
• 3. Results
• 3.1. Datasets
• 3.2. Computational experiment setup
• 3.3. SalmonTE guarantees a reliable TE expression estimation
• 3.4. SalmonTE shows a better scalability in the speed benchmark dataset
• 3.5. Discover differentially expressed TEs in ALS cell line
• 4. Conclusion
• Acknowledgments
• References
• Causal inference on electronic health records to assess blood pressure treatment targets: An application of the parametric g formula
• 1. Introduction
• 1.1. Global Burden of Hypertension
• 1.2. Challenges in Previous Efforts to Discover Optimal Target Blood Pressures
• 1.3. Causal Inference from Electronic Health Records As a Tool to Answer Difficult Clinical Questions
• 2. Methods
• 2.1. Data Acquisition from the Mount Sinai Hospital EHR
• 2.2. Problem setup
• 2.3. Parametric g formula
• 3. Results
• 3.1. Electronic Health Records Data
• 3.2. Survival time by goal blood pressure target
• 4. Conclusion
• References
• Data-driven advice for applying machine learning to bioinformatics problems
• 1. Introduction
• 2. Methods
• 3. Results
• 3.1. Algorithm Performance
• 3.2. Effect of Tuning and Model Selection
• 3.3. Algorithm Coverage
• 4. Discussion and Conclusions
• 5. Acknowledgments
• References
• Improving the explainability of Random Forest classifier - user centered approach
• 1. Introduction, Background and Motivation
• 1.1 Random Forest (RF) Classifiers
• 1.2 Related work on Explainability for Random Forest Classifiers
• 1.3 User-Centered Approach in Enhancing Random Forest Explainability - RFEX.
• 2. Case Study: RFEX Applied to Stanford FEATURE data.