General

Mibianto is a metagenomics web server designed for compositional data analysis. The web server implements a MinHash based data preparation step on user side via webassembly to compress information. This allows it to accept larger datasets as input, since the overall amount of transmitted data is reduced by around 100 folds. The compressed data is send to the server and profiled. On the results page a wide range of additional data exploration may be performed

Opening Page

The opening provides three main functionalities.
At the top of the page the ID of an existing job may be entered to access results. By default, we already provide an ID to a larger example study. Note, your result pages may also be bookmarked to access at later point in time.
At the center, a forward to the submission page is depiceted which will be further explained below.
Lastly, we provide two small .fastq files to serve as example files. Data is compressed and first requires unzipping.

Data Submission

Data submission happens in two steps.

First, all relevant metagenomic sequencing files have to be dragged/uploaded on the submission site center. Accepted data formats are .fastq and .fastq.gz. The implemented processing method is very flexible with experiment design as whole metagenome sequencing experiments with long, single-end, and paired-end reads are valid inputs. Note, the file must contain reads >50bp

Second, a sample sheet in excel (.xslx) format, tab separated value (.tsv), or comma separated value (.csv) can optionally be uploaded. Here, the Sample and File columns are mandatory as elaborated in the figure. One sample can consist of multiple files. For example if one sample consists of a paired-end experiment, two rows should be included in the file, each specifying one end. Naturally, each file can match to at most one sample. Aside from the two mentioned columns additional columns may be specified as metadata. This metadata can later be accessed during data exploration. Computed taxonomic counts and metadata will be tied to the jobid and not be hidden behind a login system. Accordingly, everyone guessing the jobid will have access to your metadata. Therefore, consider the following piece of advice.

We strongly discourage uploading of metadata that leads to de-anonymization of your data.

Taxonomic Profiling

Taxonomic profiling is performed in a snakemake pipeline on the server using the latest iteration of the GTDB database, following the recommended workflow of the developers. Note that this may take a while, especially if a large number of samples has to be processed. Accordingly, users may save the jobid or bookmark the page and return back later. After successful computation, users will automatically be forwarded fom the waiting queue to the result page.

Interactive Data Exploration

The result page consists of an interactive interface, where a wide range of different analysis can be performed. Note that each individual page has additional help information at the bottom of the page.

Citation

As of yet, Mibianto has not been published. Wish us luck!

This website is free and open to all users and there is no login requirement.