Rumming, Madis: Metadata-driven computational (meta)genomics. A practical machine learning approach. 2018
Inhalt
- Introduction
- Metagenomics, an extension to traditional ecology
- Metagenome studies and practical implications for our every day life
- Setup of a metagenomic experiment
- Finding the missing links
- MetaStone – Foundation for Metagenomic Storage of novel entities
- PP – PhenoPointer
- Machine learning-based classification
- Unsupervised learning
- Supervised learning
- Evaluation of machine learners
- Recap and comparison of ML methods
- PhenoPointer – Principles and Implementation
- Features and classification targets
- Strict classification models
- Cross-validation workflow
- Extensions to the MetaStone code base
- Final phenotype prediction models
- PhenoPointer – 13 classifiers for phenotype prediction
- Biotic Relationships
- Cell Shape
- Cell Arrangement
- Energy Source
- Gram Staining
- Sporulation
- Metabolism
- Motility
- Oxygen Requirement
- Phenotype
- Salinity
- Temperature Range
- Diseases
- Runtime and memory consumption
- PhenoPointer – Application on a real world data set and comparison to a competitor
- MVIZ – Metagenome VIZualition
- MVIZ – Principles and Implementation
- Metadata Enrichment of metagenomic community profiles
- WebUI for interactive Visualization
- How to: MVIZ
- MVIZ – Metadata-enrichment of a community profile
- From genotype over phenotype to function-driven metagenomics
- Appendix