MEG Software and Analysis: Difference between revisions
No edit summary |
No edit summary |
||
(25 intermediate revisions by 5 users not shown) | |||
Line 1: | Line 1: | ||
==== MEG Data Analysis ==== |
|||
⚫ | |||
The current version of the '''CTF DataEditor tool''' runs under CentOS (Linux) version 6.9. To support the MEG user community, |
|||
[mailto:holroydt@mail.nih.gov Tom Holroyd] has created a [https://singularity.lbl.gov/index.html Singularity] container for CentOS 6.9 and populated the container with the necessary libraries to run the DataEditor tool. This allows MEG users to run the DataEditor tool under the operating system (Windows, Mac OSX, Linux) of their choice. |
|||
⚫ | |||
Singularity is a |
|||
[https://en.wikipedia.org/wiki/Operating-system-level_virtualization Operating-system-level virtualization] solution where an operating system |
|||
can host another operating system in an isolated container. To run a Singularity container on your operating system, you need to install the Singularity software for your operating system. There are install packages for the major operating systems from the [https://singularity.lbl.gov/index.html Singularity website.] |
|||
---- |
|||
; Singularity container running CentOS 6.7 with the CTF DataEditor tool |
|||
: [https://megcore.nih.gov/MEG/ctf-6.1.14-beta.img Singularity container ctf-6.1.14-beta.img download] |
|||
: {| width=70% style="border: 1px solid black;" |
|||
|<u>Usage</u>: singularity shell --bind /data:/mnt/data ctf-6.1.14-beta.img |
|||
|- |
|||
| where the ''/data'' file system holds your CTF data set and mounts under /mnt/data inside the container |
|||
|- |
|||
| (your /home directory and the /tmp file systems are automatically visible inside the container) |
|||
|} |
|||
---- |
|||
; felix.nih.gov users |
|||
: ssh -X login@felix.nih.gov |
|||
: Copy the ctf-6.1.14-beta.img to your /data directory. |
|||
⚫ | |||
: Either add '''module load singularity''' to your ~/.bashrc startup file or type this on the command line. |
|||
: Use the following bash script to shell into the singularity container. |
|||
* [[Pyctf|Accessing CTF Datasets from Python]] |
|||
<syntaxhighlight lang="bash"> |
|||
#!/bin/bash |
|||
# goctf.sh |
|||
⚫ | |||
singularity shell --bind /data/username:/mnt/data ctf-6.1.14-beta.img |
|||
killall -s msgd 2>/dev/null |
|||
</syntaxhighlight> |
|||
⚫ | |||
The DataEditor communicates with other CTF tools using the '''msgd''' service and the DataEditor will automatically |
|||
launch this service if it isn't running. The trouble on felix.nih.gov is this service is not killed when you exit from the singularity container and you will end up with a persistent process on felix. To prevent this from occurring, make sure msgd goes away. |
|||
⚫ | |||
You may want to add this stanza to your ~/.bashrc file on felix.nih.gov. |
|||
This sets the command prompt and the command history format for inside the container |
|||
⚫ | |||
<syntaxhighlight lang="bash"> |
|||
if [ ! -z "$SINGULARITY_CONTAINER" ]; then |
|||
## PS1 is set inside the container |
|||
HISTFILE=~/.bash_history |
|||
HISTTIMEFORMAT='%Y%m%d/%H:%M:%S ' |
|||
else |
|||
export PS1='[\h \w]$ ' |
|||
fi |
|||
</syntaxhighlight> |
|||
* [[External MEG Analysis Toolboxes | Other Software packages]] |
|||
* Converting Brainsight Localizers [https://github.com/nih-megcore/nih_to_mne to tags or MNE transforms] |
|||
⚫ | |||
* Connectivity Analysis [[Connectivity Resources | Resources]] |
|||
* [[CTF DataEditor|CTF Data Editor]] |
|||
==== Basic Tutorials for New/Inexperienced Researchers==== |
|||
* [[Dataset and Task Utilities|Dataset and Task Utilities]] |
|||
Most MEG Core scripts are written in bash, which is the command line interface to Linux. Python is a more powerful programming language, and Tom and Jeff have written some of our scripts in Python, but you don't really need to know how they work, just what they do. The AFNI tutorial below is geared towards fMRI/BOLD analysis. For MEG analysis you can ignore all the BOLD things, we just use it for statistics and display of the final source reconstruction results. If you want to go more in depth, Tom can explain the ones we use. Biowulf can be employed later when you want it all to go faster. |
|||
⚫ | |||
* HPC: Introduction to Linux -- https://hpc.nih.gov/training/handouts/Introduction_to_Linux.pdf |
|||
⚫ | |||
* HPC: Bash Class -- https://hpc.nih.gov/training/bash_class/ |
|||
* Python Tutorial -- https://docs.python.org/3/tutorial/ |
|||
* AFNI Introduction -- https://andysbrainbook.readthedocs.io/en/latest/AFNI/AFNI_Short_Course/AFNI_fMRI_Intro.html |
|||
* HPC: Introduction to Biowulf -- https://hpc.nih.gov/training/intro_biowulf |
|||
⚫ | |||
==== Stimulus Presentation Software ==== |
|||
⚫ | |||
'''PsychoPy: Psychology software in Python''' |
|||
[http://www.psychopy.org PsychoPy] is an open-source application that allows you to run a wide range of neuroscience, psychology and psychophysics experiments. It’s a free, powerful alternative to Presentation™ or to e-Prime™, written in Python (a free alternative to Matlab™ ). |
|||
*[[Image:pdf.png]] [[Media:PsychoPyManual.pdf| Pyschopy documentation for release 1.90.2]] |
|||
'''Presentation: NeuroBehavioral Systems (NBS), Inc.''' |
|||
Presentation® is a stimulus delivery and experiment control program for neuroscience written for Microsoft Windows. |
|||
* [https://www.neurobs.com/presentation/docs/index_html Presentation online documentation] |
|||
'''E-prime 3: Psychology Software Tools''' |
|||
E-Prime® 3.0 software for behavioral research. ''Build your own experiments using E-Prime’s easy-to-use graphical interface. Design, collect, and analyze data – all within a few hours!'' |
|||
* [https://pstnet.com/products/e-prime/ E-Prime 3.0 website] |
|||
====Miscellaneous Documentation==== |
|||
* [[Image:adobe-ps.png]] [[Media:SensLayout-275.ps|SensLayout-275]] — a color picture showing the sensor names and relative locations (ps). |
|||
* [[External MEG Analysis Toolboxes | External MEG Analysis Toolboxes]] |
|||
* [[Image:pdf.png]] [[Media:SensLayout-275.pdf|SensLayout-275]] - a color picture showing the sensor names and relative locations (pdf). |
|||
* [[Image:pdf.png]] [[Media:FileFormats.pdf|File Formats]] - CTF MEG data file format |
Revision as of 10:35, 16 December 2021
MEG Data Analysis
This section covers all aspects of MEG data analysis. The following pages assume that you have AFNI installed and have a reasonably good idea of how to use it.
- Converting Brainsight Localizers to tags or MNE transforms
- Connectivity Analysis Resources
Basic Tutorials for New/Inexperienced Researchers
Most MEG Core scripts are written in bash, which is the command line interface to Linux. Python is a more powerful programming language, and Tom and Jeff have written some of our scripts in Python, but you don't really need to know how they work, just what they do. The AFNI tutorial below is geared towards fMRI/BOLD analysis. For MEG analysis you can ignore all the BOLD things, we just use it for statistics and display of the final source reconstruction results. If you want to go more in depth, Tom can explain the ones we use. Biowulf can be employed later when you want it all to go faster.
- HPC: Introduction to Linux -- https://hpc.nih.gov/training/handouts/Introduction_to_Linux.pdf
- HPC: Bash Class -- https://hpc.nih.gov/training/bash_class/
- Python Tutorial -- https://docs.python.org/3/tutorial/
- AFNI Introduction -- https://andysbrainbook.readthedocs.io/en/latest/AFNI/AFNI_Short_Course/AFNI_fMRI_Intro.html
- HPC: Introduction to Biowulf -- https://hpc.nih.gov/training/intro_biowulf
Stimulus Presentation Software
PsychoPy: Psychology software in Python PsychoPy is an open-source application that allows you to run a wide range of neuroscience, psychology and psychophysics experiments. It’s a free, powerful alternative to Presentation™ or to e-Prime™, written in Python (a free alternative to Matlab™ ).
Presentation: NeuroBehavioral Systems (NBS), Inc. Presentation® is a stimulus delivery and experiment control program for neuroscience written for Microsoft Windows.
E-prime 3: Psychology Software Tools E-Prime® 3.0 software for behavioral research. Build your own experiments using E-Prime’s easy-to-use graphical interface. Design, collect, and analyze data – all within a few hours!
Miscellaneous Documentation
SensLayout-275 — a color picture showing the sensor names and relative locations (ps).
SensLayout-275 - a color picture showing the sensor names and relative locations (pdf).
File Formats - CTF MEG data file format