Carlo Berg and Anders Andersson
An R Shiny App dashboard to display metagenomic data from various samples of the Baltic Sea. An external sample can be uploaded and used to predict environmental parameters from its metagenomic gene abundance data and via application of a random forest model trained with the Baltic Sea metagenomic data.
From the panel on the left side of the app several actions can be performed. First, the samples can be selected, filtered, and an external data file uploaded that is used alongside with the Baltic Sea metagenome samples. The external data file should be in tab-delimited format. Then, you can view the metagenomic data in a heatmap and view or download the count data as a table. In the final step, the external data file can be used to predict values for several environmental parameters by application of a random forest model that was trained with the Baltic Sea metagenomic data.
TPM-mormalized count data of KEGG modules or eggNOG COGs are used in a tab-delimited format. Files have to reside in the data folder. The table is formatted in wide format, i.e. one column for each sample, and one row for each functional category (KEGG module or eggNOG). The first column lists the KEGG/eggNOG identifiers (M00001, M00002 etc.). Note, that the data and random forest objects currently are not included in this repository but you can use the app in a live version (see below) where this is the case.
The external data file should be in the same format as described above and can be uploaded under the settings tab.
Tested with R 3.5.0. Packages needed: shiny, shinydashboard, tidyverse, magrittr, reshape, plotly, d3heatmap, DT, RColorBrewer, randomForest
A deployed live version is also available at cberg.shinyapps.io/baltic-sea-metagenome-dashboard/
The BONUS BLUEPRINT project has received funding from BONUS (Art 185), funded jointly by the EU and the national funding institutions of Denmark, Sweden, Germany, Finland, and Estonia.