This project contains the complete set of outputs of several different afni_proc.py (AP) runs, demonstrating various processing features and considerations. More details about the analyses and AFNI's afni_proc.py are provided in: > [**Processing, evaluating and understanding FMRI data with afni_proc.py.**][1] > > by Richard C. Reynolds, Daniel R. Glen, Gang Chen, Ziad S. Saad, > Robert W. Cox, Paul A. Taylor. This draft describes both task and resting state FMRI processing, demonstrated on one subject each. Multiple scripts are applied for each case, applying variations of potential interest. The details of the scripts and processing choices are detailed in the paper. Here, the various outputs for task-based FMRI processing are contained within the `data_task_processed` folder, and those for resting state are in the `data_rest_processed` folder. Each `data_*.tgz` file is a compressed package of the output of running a single AP command. Each can be opened by running `tar -xf NAME.tgz`. --------------------- ### Contents of the output directories Each directory contains the output of running afni_proc.py for a single subject. This includes the following five outputs in each case (here, for a participant with subject ID `sub-005`): * **ap.cmd.sub-005** : a text file containing a copy of the AP command itself, which will create a full single-subject processing script (`proc.*`, below) * **output.ap.cmd.sub-005** : the log of running the AP command, containing any warnings or errors, as well as a copy+pastable line for executing the processing * **proc.sub-005** : the full processing script created by AP, which is typically several hundred lines long; it is commented, organized by processing block and searchable * **output.proc.sub-005** : a log file from running the proc script, saving all the text that appears in the terminal * **sub-005.results** : the AP results directory, containing both intermediate and final datasets, as well as a QC HTML report and more. Altogether, these outputs help users understand their processing choices and the mechanisms by which they are carried out. All processing steps are contained in the proc script, removing any guesswork about provenance of outputs. Please see the above paper for more details. --------------------- ### Demo data and procesing scripts To get the raw data that served as input for each of these files in this demo, copy+paste the following into a terminal (or copy+paste each of the links into a browser tab): curl -O https://afni.nimh.nih.gov/pub/dist/tgz/demo_apaper_afni_proc_rest.tgz curl -O https://afni.nimh.nih.gov/pub/dist/tgz/demo_apaper_afni_proc_task.tgz Then to unpack it, copy+paste this in a terminal: tar -xf demo_apaper_afni_proc_rest.tgz tar -xf demo_apaper_afni_proc_task.tgz Each demo folder contains its own set of unprocessed data, pre-afni_proc.py processed data (like FreeSurfer, sswarper2 and/or timing_tool.py), and a scripts directory. A complete description of each of the resting state and task demo data and scripts are provided [here][2]. The GitHub repository for the set of both HPC and desktop scripts is: https://github.com/afni/apaper_afni_proc [1]: https://arxiv.org/abs/2406.05248 [2]: https://github.com/afni/apaper_afni_proc?tab=readme-ov-file#apaper_afni_proc