Inhalt

•  
   Titlepage
•  
   Zusammenfassung
•  
   Abstract
•  
   Acknowledgements
•  
  1 Introduction
•  
   1.1 Motivation
•  
   1.2 Metagenomics
•  
   1.3 Single-cell genomics
•  
   1.4 Flow cytometry
•  
   1.5 Structural overview
•  
  2 Automatic discovery of metagenomic structure
•  
   2.1 Background
•  
  2.2 Methodology
•  
   2.2.1 Data representation
•  
  2.2.2 Dimensionality reduction
•  
   Problems of high-dimensional spaces
•  
   t-SNE
•  
   2.2.3 Cluster Analysis
•  
   Clustering algorithms
•  
   Clustering evaluation
•  
   2.2.4 Binning pipeline
•  
   2.3 Evaluation
•  
   2.3.1 Data
•  
   2.3.2 Dimensionality reduction
•  
   2.3.3 Data representation by oligonucleotide frequencies
•  
   2.3.4 Clustering algorithms and cluster validation
•  
   2.3.5 Application to complex metagenomes
•  
   2.4 Summary
•  
  3 Single-cell genome contamination detection
•  
   3.1 Background
•  
   3.2 Methodology
•  
  3.2.1 Reference-free detection
•  
   Steps 1&2: Data pre-processing and dimensionality reduction
•  
   Step 3: Estimating contamination confidences
•  
  3.2.2 Reference-based detection
•  
   Step 4: Sequence classification
•  
   Step 5: 16S rRNA gene prediction
•  
  3.2.3 Decontamination
•  
   Step 6: Manual cleansing
•  
   Step 7: Taxonomy annotation and automatic cleansing
•  
   Step 8: Result visualization
•  
  3.3 Results
•  
   3.3.1 Computational performance
•  
   3.3.2 Evaluation data sets
•  
   3.3.3 Supervised analysis
•  
   3.3.4 Unsupervised analysis
•  
   3.3.5 Assessing the purity of metagenome bins
•  
   3.4 Discussion
•  
   3.4.1 Influence of assembly size and quality
•  
   3.4.2 Influence of horizontal gene transfer and repeats
•  
   3.5 Summary
•  
  4 Identification of flow cytometry cell populations
•  
   4.1 Background
•  
  4.2 Methodology
•  
   4.2.1 Pipeline
•  
   1: Input FCM data
•  
   2: Density estimation
•  
   3: Example gates
•  
   4: Alignment and gate transfer
•  
   4.2.2 Evaluation measures
•  
   4.2.3 Quality checking cell population thresholds
•  
   4.2.4 Implementation and computational complexity
•  
  4.3 Study design and evaluation data sets
•  
   4.3.1 Mice data
•  
   4.3.2 FlowCAP data
•  
  4.4 Results
•  
   4.4.1 Mice data
•  
   4.4.2 FlowCAP data
•  
   4.4.3 Runtime
•  
   4.4.4 Comparison to nearest-neighbor gating
•  
   4.4.5 Comparison to DeepCyTOF and FlowSOM
•  
   4.5 Discussion
•  
   4.6 Summary
•  
   5 Conclusion
•  
  A Appendix
•  
  A.1 Software Availability
•  
   A.1.1 Acdc
•  
   A.1.2 FlowLearn
•  
   A.2 List of genomes used in the NCBI-9 data set
•  
   A.3 Results on the CAMI data
•  
   A.4 Acdc parameters
•  
   A.5 Evaluation of the optimal number of nearest neighbors m
•  
   A.6 Description of the simulated data set
•  
   A.7 Description of the mix data set
•  
   A.8 Results on the FlowCAP data set
•  
   Bibliography
•  
   Colophon