Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: THEend8_
Visual Analysis of the Ocean Microbiome
Background
Data visualization has become an important tool for explorative data analysis as well as for presentation and communication of data in many application domains. A domain that has become increasingly data driven over the last decades are biosciences, and in particular when it comes to studies of the microbiome and other genome sequenced data. In this summative assessment, you are asked to design and implement an interactive multiple coordinated views visualization that support analysis of data from a study of the ocean microbiome, using different visualization methods.
The focus of the tasks in the assessment is on visualization of heterogeneous and multivariate (high dimensional) data, interactive visualization and multiple views, heuristic evaluation, and visualization of uncertainty.
Data context
The oceans are the largest cohesive eco-system on earth, and a greater understanding of this eco-system is important for the preservation of the planet as well as for understanding of how organisms have evolved since life began. The data that you will work with originates from a two-and-a-half-year expedition with the schooner Tara, during which oceanic samples were collected from 210 stations across the world oceans.
User context
The end user of the visualization that you will develop would typically be a microbiologist or another domain expert in a bioscience field. The aim of their analysis would be to increase their knowledge of the ocean microbiome, and analysis questions of particular interest may for example include:
? Which microbes are detected at the highest levels overall in the oceans?
? Which microbes are detected at the highest levels in certain regions of the oceans?
? Are there differences in microbe detection levels that can be linked to other features of the oceanic samples, for example the geographic region, sample depth etc?
? Are there differences between taxonomic levels, which can be linked to other features of the oceanic samples?
The data
You will be provided with a set of different spreadsheets to work with, which have gone through some initial formatting and cleaning. The full dataset include data related to 135 samples that were taken from different oceanic regions.
The detection levels of 35,650 Operational Taxonomic Units (OTUs) were recorded for the individual samples. Detection levels are sometimes referred to as the abundance of the OTU. OTUs are close approximations of microbial species, which are extracted through clustering of DNA sequences, so you can think of an OTU as being the same as a microbial species (such as a bacterium) . The OTUs also have an associated hierarchical taxonomy through the biological classification system (https://en.wikipedia.org/wiki/Taxonomy_(biology)), and are often converted into higher levels in the taxonomy for analysis, since an OTU name generally has no biological meaning. Analysis is quite often carried out and reported at Genus level.
In addition to the OTU detection levels, there area range of contextual data associated with the samples (i.e. metadata) . From a data science and visualization perspective, the OTUs are generally treated as data variables (dimensions) and the samples are data items.
You will be provided with the following datasets, in comma separated file format (csv):
? Tara_OTUtableTax_full.csv: Each row in this file corresponds to a unique OTU (microbial species). The first six columns include the taxonomic classification for each OTU at the following hierarchical levels: Domain, Phylum, Class, Order, Family, Genus. The original taxonomic classification of the OTUs included a lot of missing values, as a result of OTUs that were not identifiable at all levels in the taxonomy. The highest level where nearly all OTUs were identified was the Class level. Due to this, the missing values have been replaced with the Class name of the OTU, followed by (undef) (i.e. a Cyanobacteria OTU is referred to as Cyanobacteria(undef) at all levels where it has not been classified). The seventh column include a unique OTU-id, which has no biological meaning. The remaining columns each correspond to a sample, with a unique sample id as heading. The cells represent the relative detection level (relative abundance) of OTUs in samples as a percentage value, thus the sum of each column is 100%.