A reporting pipeline for visualization of immunopeptidomics MS data.
MhcVizPipe is intended for rapid assessment and comparison of general sample quality and composition across a single or multiple files. Example use cases of MhcVizPipe are assessment of QC samples or comparison of samples during method development.
MhcVizPipe connects the bioinformatic tools NetMHCpan, NetMHCIIpan and GibbsCluster and generates organized and easy-to-understand reports in HTML format. The reports are fully portable and can be viewed on any computer with a modern web browser.
Here are some of the features of MhcVizPipe:
The "Sample Overview" sections gives you a quick look at the peptide length distributions and the overlapping sets of peptides between multiple samples.
The "Annotation Results" section summarizes the results of NetMHCpan(4.0 or 4.1) or NetMHCIIpan4.0. A table shows the number and proportion of strong, weak and non-binders per sample per allele while overall binding characteristics of each sample are displayed in a single figure.
The "Sequence Motifs" section highlights the prominent motifs identified by GibbsCluster, including motifs generated from the analysis of all peptides ("Unsupervised GibbsCluster") and motifs generated for each allele-specific peptide list as identified by NetMHCpan or NetMHCIIpan ("Allele-specific GibbsCluster").
MhcVizPipe is intended for situations where you want a quick and consolidated view of the general quality and composition of one or more immunopeptidomics samples:
- QC samples
- Compare numbers of binders, distribution of binders per allele and prominent binding motifs with previous QC sample to check consistency of data
- Quality check of a batch of data returned from a core facility
- Rapidly check for overall sample quality
- Check for possible labeling issues (e.g. switch-ups of class I and class II)
- Comparison of multiple samples in a method development experiment
- Compare general composition across multiple conditions
MhcVizPipe is not intended for situations where you want data mining of your sample or if the sample is uncharacterized. For example:
- Finding new motifs in data
- MhcVizPipe is designed to show you the most prominent motifs in your data, not find every single possible motif. If you are looking for something that might be hidden or requires manipulation of the data to find, you will need to do this analysis yourself.
- Uncharacterized samples
- MhcVizPipe expects you to know your sample(s). E.g. you should know if it is class I or class II and what alleles are expressed.
- As a replacement for a bioinformatician
- MhcVizPipe automates some routine and time-intensive tasks for you, but it does provide biological insight beyond what you can see in the report! MhcVizPipe gives you a general characterization of the sample, not a deep biological interpretation.
MhcVizPipe will run natively on Linux (e.g. Ubuntu) or MacOS systems. It can also be installed on Windows 10 systems using the Windows Subsystem.
There are no hard memory or CPU requirements for MhcVizPipe. However, because MhcVizPipe utilizes multi-threading you will see performance increases on systems with higher numbers of CPUs.
Below you will find a brief overview of the installation steps and usage of the tool. For more details please visit the wiki.
Note that MVP runs natively on Linux and Mac systems. If you have a Windows computer, you can follow these instructions to install it using the Windows Subsystem for Linux.
There are two ways to install MhcVizPipe:
- From the releases section, download a standalone package
which contains Python and all the required Python packages. The only thing
you need to do is download GibbsCluster, NetMHCpan and NetMHCIIpan and extract them into a directory called
"tools"
inside the MhcVizPipe directory. This option is completely portable, e.g. you could move the program onto
a USB stick and use it on any compatible system.
- Because of system dependencies of the standalone Python distribution we are using, we cannot guarantee the portable package will work on every computer. However, we have tested in Ubuntu 16.04-20.04, LinuxMint 18, several versions of Mac OSX and using the Windows Subsystem for Linux, and generally have had no problems. If you encounter problems, please try option 2 below.
- Install MhcVizPipe into an existing Python environment using
pip install MhcVizPipe
and configure the settings to point to existing installations of GibbsCluster, NetMHCpan and netMHCIIpan.
For complete instructions, visit the wiki.
From any terminal, enter the following command:
MhcVizPipe
You should see the following message:
========================================
MhcVizPipe v0.3.1
Welcome to MhcVizPipe! To open the GUI, open the following link
in your web browser: http://0.0.0.0:8080
(To do this, most likely right click the link and choose something
like "Open URL" or "Open in browser". If that doesn't work,
copy and paste it into your browser.)
If this is you very first time running MVP, when you open the
link a window should appear which will help you install GibbsCluster,
NetMHCpan and NetMHCIIpan if you have not already.
For a brief introduction to using the GUI, click the link to
"help and resources" near the top of the GUI. For more information
and the latest updates please visit our GitHub repository:
https://github.com/kevinkovalchik/MhcVizPipe.
========================================
You will notice this link in the message: http://0.0.0.0:8080
. Right click and select "Open in browser" and
the MhcVizPipe will open up in your web browser.
For detailed usage, see the wiki usage page.
If you need further help please open an issue!